Powered by massive leaps in AI, deep learning, and on-device processing, today’s automatic speech recognition (ASR) systems are faster, more accurate, and more context-aware than ever before. Far from simply converting voice to text, speech AI systems can help shorten your workday, secure your home, and even drive your car.

Key takeaways

• Automatic speech recognition (ASR) uses technology to convert speech into text. Speech AI systems use AI and NLP models to extract insights from this data and make decisions.
• Digital assistants like Siri and Alexa use speech recognition to interpret voice commands and act as agents on users’ behalf.
• Speech recognition technology is used in a variety of industries to improve efficiency, safety, and outcomes. For example, virtual meetings are transcribed and summarized instantly. And while driving, people can use voice commands to monitor their navigation.

What is speech recognition technology?

The umbrella term ‘speech AI’ encompasses several distinct, yet interconnected, technologies, all powered by machine learning to analyze audio data.

Automatic speech recognition (ASR) is the technology that converts spoken words or recorded audio into written text. This is also referred to as ‘speech-to-text.’ ASR’s primary goal is accurate transcription.¹

Examples of ASR include:

• Using your smartphone’s dictation feature to record a voice note that is automatically transcribed and sent as a text message.
• A doctor using a dictation system to transcribe patient notes directly into an electronic health record.
• Live captioning on a TV broadcast or online video.

Natural language processing (NLP) is a subset of machine learning in which computers interpret, manipulate, and comprehend human language. Speech recognition is an example of NLP for audio data. There is also text-based NLP that analyzes large corpuses of written information.²

Examples of NLP include:

• A language translation app that converts text from one language to another.
• Automatic grammar and spell-check systems.
• Email filters that automatically detect spam based on the language used and metadata.

Audio intelligence refers to the application of machine learning models to extract insights from audio data and complete tasks like sentiment analysis or content moderation. Audio intelligence relies on automatic speech recognition to convert audio data into digital information and NLP models to complete tasks.³

Examples of audio intelligence include:

• A smart home device detects the sound of a fire alarm or a window breaking to send a security alert to authorities.
• Software that listens to recorded customer service calls and automatically identifies and categorizes the customer’s tone as positive or negative.

What are speech recognition digital assistants?

Digital assistants are designed to help people perform or complete basic tasks and respond to queries. With the ability to access information from vast databases and various digital sources, these robots help to solve problems in real time, enhancing the user experience and human productivity.

Popular digital assistants, include:

• Amazon’s Alexa
• Apple’s Siri
• Google’s Google Assistant
• Microsoft’s Cortana

Five applications of speech recognition technology for 2026

Speech recognition technology and the use of digital assistants quickly moved from mobile phones to homes. Today, its application is apparent across crucial industries, including healthcare, banking, and marketing.

1. Speech recognition in the workplace

Speech recognition technology in the workplace is moving beyond simple voice commands to become a primary engineer for productivity and efficiency. The core goal of this technology in the workplace is the elimination of low-value, repetitive administrative tasks, freeing employees to focus on strategic work.

• Meeting and notetaking automation: AI serves as a virtual meeting scribe for the hybrid work environment. Using ASR technology, platforms can automatically transcribe conversations in real-time, even with multiple speakers or diverse accents. These systems immediately extract insights, generate summaries, and identify next steps or action items to share with participants.⁴

• Customer service: Speech AI and ASR are leveraged in the customer service industry to augment human work. For example, real-time agent assist features can do live sentiment scoring and summary of customer calls, providing immediate, actionable insights for representatives.⁵

• Workplace accessibility: Speech recognition is critical for making in-person and hybrid work environments more accessible for all employees. Real-time transcriptions provide support for people with hearing impairments or language barriers.

2. Speech recognition in banking

For the banking and financial services industry, Speech AI can help achieve two main goals: enhancing security and fraud prevention and creating a frictionless customer experience. The technology moves beyond simple account inquiries to handle complex authentication and compliance requirements.

• Personalized self-service: Speech-based tools can allow users to schedule recurring payments, check available funds, and review past transactions over the phone. For example, some bank mobile apps offer users the ability to use their microphone to send money via Zelle or transfer funds.⁶

• Call routing: Speech AI can interpret live phone calls to immediately route customers to the right bank department or specialist, reducing the need for transfers. This can also lead to shortened resolution times.⁷

3. Speech recognition in marketing

Speech AI has added a new dimension to how marketers interact with consumers, making search and shopping more conversational and immediate. This shift requires marketers to pivot their digital strategies to focus on how people talk, not just how they type.

• Voice commerce: Shopping via voice command is expected to produce $81.8 billion in sales worldwide in 2025.⁸ Consumers use voice assistants to research products, check prices, and make purchases. This move towards V-commerce mandates that brands optimize product listings and checkout experiences to work via voice commands.

• Conversational SEO: Voice queries tend to be longer, more conversational, and more question-based than typed queries. Marketers must optimize content for long-tail keywords and answer-focused content.⁹

4. Speech recognition in healthcare

In healthcare settings — where accuracy, speed, and hands-free operation are matters of patient safety — Speech AI offers transformative potential for clinical efficiency and reducing physician burnout.

• Documentation: Digital scribes use ASR to document provider-patient interactions and then summarize the visit, populate diagnostic fields, and create billing codes. In a study of nurses who used speech recognition systems to record and document nursing reports, researchers found that paperwork reduction, performance improvement, and cost reduction were some of the most common benefits.¹⁰

The most significant concern using speech recognition in healthcare is the content the digital assistant or AI platforms can access. Hallucinations, transcription errors, and omissions all pose risks to patient privacy and safety. Proper guardrails and oversight can help mitigate these risks.¹¹

5. Speech recognition and the Internet of Things

Speech recognition is a core component of the Internet of Things (IoT), acting as the interface between interconnected smart devices. This is expanding beyond basic smart home features into complex multimodal systems and large-scale industrial applications.

• Automotive control: In vehicles, Speech AI can help manage navigation, climate control, and infotainment systems. By allowing drivers to interact using natural speech, the technology reduces cognitive load and visual distraction.¹² Researchers are also experimenting with ‘hearing cars’ — vehicles equipped with external microphones and AI to help detect and classify hazards that autonomous cars can’t see. Approaching emergency vehicles are the first hazard being tested, but future capabilities could include sensing pedestrians or failing brakes.¹³

• Multimodal security: The future of IoT is multimodal, combining voice with other inputs like computer vision (for facial recognition) and gesture control. For example, a system can confirm a user’s identity via voice biometrics while simultaneously verifying their face, offering a higher level of security for unlocking doors or authorizing sensitive transactions.

Future applications

The speech and voice recognition market is experiencing an explosion of growth, driven by breakthroughs in AI and the rapid integration of large language models (LLMs). The global conversational AI market alone is projected to reach over $136 billion by 2035.¹⁴

The future of speech recognition could include developments in contextual intelligence and multimodal integration.

• Context-aware AI: Autonomous AI agents are systems that can set goals, plan complex tasks, and act with little human intervention. This could mean that instead of waiting for a command, future voice assistants might be able to anticipate user needs. For example, automatically adjusting the vehicle temperature based on a passenger talking about how warm they are.

• Multimodal integration: When Speech AI is integrated with visuals, gestures, and sensor data, it will become more powerful. For example, visual analysis of speakers’ lip movements could help reduce transcription errors. Another example of a future use case is with large-language model (LLMs) applications, like Gemini Live. Users can speak directly to an LLM and have a conversation with the AI system about uploaded files, photos, or a live feed from the phone camera.¹⁵

Explore online artificial intelligence courses and machine learning courses to explore speech AI and the models that power these platforms.

MIT Sloan School of Management
Artificial Intelligence: Implications for Business Strategy

Saïd Business School, University of Oxford
Oxford Artificial Intelligence Programme

MIT Sloan School of Management
Artificial Intelligence in Health Care

MIT Sloan School of Management
Artificial Intelligence in Pharma and Biotech

The post Applications of speech recognition appeared first on GetSmarter Blog.

While these topics may seem theoretical, their impact is deeply practical. From boosting performance to enabling scalability, they play a critical role in creating systems that are smart, responsive, and built to last.

What are data structures?

Data structures are methods of organizing and storing data to enable efficient access and modification.¹ Choosing the right structure helps reduce complexity and improve performance and scalability.

Common data structures include:²

• Arrays — Fixed-size data sets with predictable access patterns, such as storing high scores or calendar months.
• Linked lists — Useful for collections where elements are frequently added or removed, like playlists or to-do lists.
• Stacks and queues — Manage data in order. Stacks (Last In, First Out) support undo features. Queues (First In, First Out) handle task scheduling and messaging.
• Trees and graphs — Represent hierarchical or networked data. Examples include organizational charts, folder structures, or social networks.
• Hash tables — Optimize lookup speeds. Commonly used in caching, indexing, and authentication systems.

Each of these structures offers unique strengths, making them essential tools in every developer’s toolkit.

What are algorithms?

Algorithms are step-by-step instructions to complete tasks. Cooking recipes and math formulas can be thought of as algorithms. They allow us, and machines, to use proven methodologies in order to solve problems. Machine algorithms are developed by programmers and work in conjunction with data structures to deliver computational results efficiently.³

Common algorithm types include:⁴

• Sorting algorithms — Arrange data in a specific order to enable faster retrieval. Popular examples include quicksort, mergesort, and bubble sort.
• Searching algorithms — Help locate specific elements within a data set. Binary and linear search are common techniques.
• Graph traversal algorithms — Explore or map networks using depth-first or breadth-first search approaches.

How to choose the right data structure and algorithm

Not all data structures and algorithms are created equal. Choosing the right one can impact everything from application speed to maintainability.

Every operation is performed at a certain computational cost or complexity. Big O notation is used to describe the time and space that an algorithm takes up. It gives programmers a way to compare the efficiency and performance of different data structures or algorithms.

For example, there is linear complexity — O(n) — in which an algorithm’s runtime grows directly in proportion with the size of the input. Every additional input increases the runtime an equal amount. In contrast, in an algorithm with logarithmic complexity (O(log n)), the runtime increases proportional to the logarithm of the input. With every additional input, the runtime grows by a smaller and smaller amount.⁵

Selecting the right data structure and algorithm can change the Big O notation of your program. Here are several practical considerations to help maximize efficiency:⁶

• Data volume — Simple arrays or lists work for small data sets, while trees or graphs are better for large-scale or interconnected data.
• Operation frequency — If your application requires frequent inserts and deletes, dynamic structures like linked lists or balanced trees are more effective.
• Access patterns — Need fast lookups? Use hash tables. Processing tasks in order? Stacks and queues are ideal.

By aligning the structure to your use case, you create systems that are more efficient and scalable.

Where do data structures and algorithms appear in the real world?

Data structures and algorithms aren’t just academic — they shape the tools and services we use every day:

• Machine learning — Structures like matrices and trees are foundational in modeling and training algorithms.
• Web applications — DOM traversal and front-end rendering rely on tree structures and stack logic.
• Everyday tools — GPS route optimization, email delivery queues, and real-time search suggestions are all powered by well-chosen structures and algorithms.

Understanding how these systems work under the hood helps developers create more resilient and responsive applications.

Why learning these concepts matters for your career

Mastering data structures and algorithms is essential for professionals in software development, AI, data science, and more. Benefits could include:

• Interview readiness — These topics are frequently tested during technical assessments.⁷
• Performance credibility — Writing optimized code signals a strong technical foundation.
• Career resilience — As automation expands, demand for algorithmic thinking and analytical skills continues to grow.⁸

By investing in these skills, you could enhance your technical reputation and long-term employability.

How to start learning data structures and algorithms

You don’t need a computer science degree to begin. Structured learning options like data science and analysis courses offered by GetSmarter provide in-depth exploration, real-world applications, and expert instruction to build confidence and credibility.

Start by choosing a language like Python, which is known for readability and simplicity. Then, explore interactive content that includes visualizations, hands-on coding tasks, and incremental challenges. Reinforcing your learning with personal projects or real-world case studies helps ensure long-term retention and application.

Frequently asked questions (FAQ)

What’s the difference between an array and a linked list?

Arrays are indexed and fixed in size. Linked lists are dynamic and support frequent changes.⁹

Are algorithms only for developers?

No. Professionals in many fields benefit from algorithmic thinking when working with complex data.

What’s the fastest sorting algorithm?

It depends on the data. Quicksort is fast on average; mergesort is more stable for sorted inputs.¹⁰

Which programming language should I use to learn data structures and algorithms?

Python is often recommended for its readability and strong support libraries. Java and C++ are also commonly used. Ultimately, the programming language you choose should align with the kinds of projects you want to create, the industry you want to join, and your prior experience with math and coding.

What’s a real-world use case for graph traversal algorithms?

They’re commonly used in social networks to identify relationships, in maps for shortest path calculations, and in AI for planning.

Conclusion: Building a smarter technical foundation

Understanding data structures and algorithms helps professionals build logical, efficient, and scalable solutions. These foundational skills are essential for navigating modern technology and maintaining career competitiveness in a rapidly evolving landscape.

Whether you’re preparing for technical interviews or scaling complex systems, mastering these building blocks of software engineering is a smart step forward.

University of Cape Town (UCT)
Data Science with Python online short course

University of Cape Town (UCT)
Data Analysis online short course

MIT Sloan School of Management
Unsupervised Machine Learning: Unlocking the Potential of Data

The post Understanding data structures and algorithms: A practical guide for future-focused technologists appeared first on GetSmarter Blog.

The world is driven by data, and it’s being analyzed every second, whether it’s through your mobile GPS, your Netflix habits, or the items sitting in your online shopping cart. Every business sector looks to data for market insights and ultimately, to generate growth and revenue. 

Choosing the right type of data analysis for your task, and the techniques and tools to go with it, is an important step. Keep reading to learn more about how data analytics is used to take massive amounts of information and make predictions and decisions.

Key takeaways

• The amount of data being generated is growing at an unprecedented rate, with 90% of the world’s data being created in just the last few years.
• There are four main types of data analysis — descriptive, diagnostic, predictive, and prescriptive — each serving a unique purpose, from understanding what happened in the past to recommending a course of action for the future.
• There are a variety of techniques to help you accomplish data analysis techniques, from data visualization and A/B testing, to machine learning tactics like NLP and decision trees.

What is data analysis?

Data analysis, or data analytics, is the process of applying logical and mathematical techniques to datasets in order to discover patterns and useful information, often to aid in decision-making.³ 

It’s often used on an industrial scale, helping organizations make calculated and informed business decisions. As the Internet of Things (IoT) expands and technology develops, new forms of data mining and analysis are constantly emerging.

Big data is characterized by the three V’s: Volume, the massive scale of data; velocity, the speed at which it’s generated and processed; and variety, the wide range of structured and unstructured data formats.⁴

Velocity is particularly important, as the need for real-time analysis has led to the integration of big data with advanced technologies like machine learning and artificial intelligence.

What are the types of data analysis?

The different types of data analysis provide a structured approach to gaining insights from data, moving from understanding what has happened to predicting what will happen and, finally, prescribing a course of action. The four main types of data analysis are:⁵

Descriptive analytics: Summarizes historical data to explain what happened. This could include generating reports, KPI dashboards, and summaries. It helps you understand the past and present state of data.

Diagnostic analytics: Focuses on historical data to determine why something happened. It involves techniques like regression analysis and A/B testing to find the variables that lead to a specific outcome.

Predictive analytics: Uses statistical models and machine learning to forecast what is likely to happen in the future. It can be used for tasks like sales forecasting, risk assessment, and identifying behavior trends.

Prescriptive analytics: Recommends what should be done to achieve a desired outcome. It leverages advanced techniques like simulation and natural language processing.

10 examples of big data analytics techniques

Big data analytics techniques function in a two-fold manner: processing data streams as they emerge and performing batch analysis on data as it accumulates to identify patterns and trends. As data generation accelerates, these techniques must evolve to handle the speed, scale, and depth of information.

1. Data mining

Data mining extracts patterns from large data sets by combining methods from statistics and machine learning, within database management. Today, it is increasingly automated and integrated with AI, allowing for more complex pattern detection.⁶

• Tools: Data-mining tools include programming languages like Python or R, and proprietary tools like KNIME and RapidMiner, which offer visual workflows and pre-built algorithms.

• Example: A retail company might use data mining to analyze customer purchasing histories and identify which segments are most likely to respond to a new product promotion.

2. Data visualization

Data visualization is the graphical representation of data and information. Through visual elements like charts, graphs, and dashboards, analysts communicate insights and help stakeholders make decisions. It can also make complex data understandable and accessible to non-technical users.⁷

• Tools: Data visualization softwares like Tableau and Power BI are powerful packages with built-in charting tools. There are also coding libraries that can power highly-customized visualizations, like D3.

• Example: A business dashboard can display real-time sales data using a line chart to show a trend over time or a heat map to show customer density.

3. Cluster analysis

Cluster analysis is a type of data mining that uses unsupervised machine learning in order to group data points into distinct clusters based on their similarities. The goal is to identify hidden groupings or structures within the data without any pre-existing labels.⁸

• Tools: Common tools and libraries for cluster analysis include Scikit-learn in Python and R.

• Example: A bank might use cluster analysis to segment customers’ behavior and detect fraudulent transactions by identifying groupings of unusual activity.

4. A/B testing

This data analysis technique involves comparing a control group with a variety of test groups to discern what changes will improve or change a given objective variable.⁹

• Tools: Software and platforms like Adobe Target, A/B Smartly, and VWO offer features for creating, monitoring, and analyzing a variety of web-based A/B tests.

• Example: A marketing team could use A/B testing to determine which website layout or ad copy leads to the highest number of conversions.

5. Regression analysis

Regression analysis is a statistical method for estimating the relationships among variables. It’s used to understand how the value of a dependent variable changes when one of the independent variables is varied.¹⁰

• Tools: Statistical software like SAS and SPSS, and programming languages like R and Python are used for regression analysis.

• Example: A real estate agent could use regression analysis to determine the relationship between house size and selling price.

6. T-tests

A t-test is a statistical hypothesis test used to determine if there is a significant difference between the averages of two groups. It helps to see if they are genuinely different or if the observed difference is due to chance.¹¹

• Tools: T-tests are available in statistical software like R and Python’s SciPy library.

• Example: A teacher could use a t-test to compare the average test scores of students who tried a new study method versus those who tried an old method, to see if there is a meaningful difference in their performance.

7. Machine learning

Machine learning automates model building for analytics. It enables computers to automatically learn from data without explicit programming to make inferences, predictions, and recommendations.¹²

• Tools: Popular tools and platforms for machine learning include Databricks, KNIME, and cloud-based services like Google Cloud AI and Amazon SageMaker.

• Example: Machine learning can be used by banks to detect fraud and identify suspicious credit card transactions automatically.

8. Time-series analysis

Time-series analysis is a statistical technique that analyzes data points collected over a period of time. The goal is to identify patterns, trends, and seasonal changes in the data.¹³

• Tools: Some tools for time-series analysis include Python libraries like Pandas, Statsmodels, and specialized platforms like Amazon Forecast.

• Example: Weather forecasters might use this kind of analysis to predict future weather conditions based on past temperature, pressure, and wind patterns.

9. Decision trees

Decision trees are a type of supervised machine learning algorithm that can be used for both classification and regression. It works like a flow chart, identifying optimal split points based on feature values to create pure subsets.¹⁴

• Tools: Decision-tree algorithms are implemented using various machine learning libraries, including Scikit-learn in Python and R.

• Example: A business could use a decision tree to predict whether or not a customer will buy a product, based on factors such as previous purchasing history, age, and location.

10. Natural language processing

Natural language processing (NLP) is a subset of AI and machine learning that uses algorithms to analyze, understand, and generate human language. Large language models (LLMs) and generative AI enable tools to process massive amounts of unstructured text data, such as emails, social media posts, and customer reviews.¹⁵

• Tools: Popular NLP tools include programming libraries like NLTK and spaCy, and the IBM Watson platform.

• Example: NLP is used for machine translation services, like Google Translate, to process input text and return that same information in a number of different languages.

Learn how to sort, analyze, and interpret data to inform business strategy with data analysis short courses on GetSmarter.

University of Cape Town (UCT)
Data Analysis online short course

THE LONDON SCHOOL OF ECONOMICS AND POLITICAL SCIENCE (LSE)
Data Analysis for Management online certificate course

Rice Glasscock
Data Analysis and Visualization online short course

THE LONDON SCHOOL OF ECONOMICS AND POLITICAL SCIENCE (LSE)
Applied Data Visualisation and Analysis for Business

The post Big data analytics techniques and types appeared first on GetSmarter Blog.

When implemented effectively, business process modeling can transform project management challenges into opportunities for shared understanding, innovation, and success.

Fundamentally, business process modeling involves identifying, defining, and representing a process in its entirety, in order to aid communication of that process. 

The visual references and diagrams used in modeling systems can help you:

• Identify tasks that are redundant and remove them
• Spot bottlenecks in the process and improve process efficiency by eliminating them
• Identically repeat processes and train new staff to do so

There are several techniques in business process modeling:

1. Business Process Modeling Notation (BPMN)

The Business Process Model and Notation (BPMN) is the global standard for modeling business processes. It is composed of a system of graphical notations used to visualize end-to-end tasks and workflows. Designed for stakeholders who manage and design business processes, BPMN emphasizes a combination of ease-of-use and specificity to meet the different needs of global organizations.¹

Strengths:²

• Individual organizations don’t need to invent their own visual system
• A standardized language helps close communication gaps between stakeholders
• Includes a library of process flows and rules to train new employees on
• Can be created using online tools like LucidChart and Visio

Limitations:³

• The highly rigid framework can make process innovation more complicated
• Challenging for first-time users to read
• May not be flexible for certain dynamic and non-sequential processes

BPMN symbols fall into the following basic categories:⁴

• Flow objects. The main elements that define a process are broken into three types of objects:
  
  • Events are displayed inside circles. These are things that “happen” during the course of a process and usually have a cause and impact.
  • Activities fit into rectangular boxes. This represents the work performed within a process.
  • Gateways are represented as diamond shapes. These are used to control how the process flows and either converges or diverges.

• Connecting objects. Show how tasks are connected and in what sequence they occur.
  
  • Sequence flows are displayed as solid lines. These depict the order of activities.
  • Message flows are displayed as dashed lines. These depict the transfer of information between participants.

• Swimlanes. Make provisions for partitioning a set of activities from others. They comprise “Pools” and “Lanes”.

• Artifacts. Provides additional information that isn’t captured in a sequence or message, but does not directly affect the process.⁵
• • Groups use dot-dash rectangles to logically relate components.
  • Text annotations are displayed inside square brackets.
  • Associations link text annotations to other elements via dotted lines.

2. Unified modeling language diagrams

Unified modeling language (UML) is a general-purpose approach to visually modeling and documenting processes. UML was initially developed by software developers, but has been successfully used in business process modeling.

It uses 14 types of diagrams to display an object-oriented approach to process visualization. This means that objects are modularized to include things like class, inheritance, and polymorphism. The types of diagrams are split into two categories: behavioral and structural, which display the dynamic and static aspects of a system, respectively.⁶

Strengths:⁷

• Popularity among software developers makes it highly recognized and interpretable for teams
• Versatile enough to model processes, applications, and software engineering projects
• Can use tools like LucidChart, Visio, and Draw.io

Limitations:

• Can have a steep learning curve to understand the different diagrams and object types
• Some aspects of the design can lead to ambiguity and different interpretations
• Diagrams can become overly complex for large systems

3. Flow chart technique

This is a generic graphic representation of a process, product, or system, which gives people involved in the project a single reference point. Flow charts use basic shapes and arrows to define relationships between steps, decisions, or data.⁸

Strengths:

• Highly flexible without prescriptive shapes and visual systems
• Intuitive and easy to understand

Limitations:

• Difficult to draw for large and complex processes
• No standard language for general flow charts

4. Data flow diagrams

Data flow diagrams (DFDs) show the flow of information through a process. Using pre-defined symbols, DFDs visualize data inputs, outputs, storage points, and routes. The symbols generally fall into four types:⁹

• External entities are represented using squares. These are outside systems that are either the source or destination of data.
• Process is shown using rounded rectangles. These are any steps that change the data and produce an output.
• Data stores are visualized using open-ended rectangles. These are files or repositories that hold information.
• Data flow is represented via arrows. This is the route that data takes between the external entities, process, and data stores.

Strengths:

• Well suited for analysis and modeling data-heavy systems
• Help manage cybersecurity risks

Limitations:

• Less applicable for interactive or real-time software
• Limited scope and does not apply to more general business processes
• Requires technical knowledge of the data flow to create

5. Role activity diagrams

Role activity diagrams (RADs) are used to map out the intangible roles or ideas of behavior that are desired within the company. These can often be functions within the business, systems in IT, or customer and supplier roles.¹⁰

Strengths: 

• Easy to read and understand
• Provide a different perspective on a process

Limitations:

• Limited to mapping roles and behavior, not all processes

6. Interaction diagrams

A component of UML, interaction diagrams are business process models that graphically illustrate the interaction of various processes with each other within a system. Interaction diagrams come in two forms: sequence diagrams and collaboration diagrams.¹¹

Strengths: 

• Helps describe the flow of messages within a system
• Great for illustrating collaborations
• Identify possible connections between lifeline elements

Limitations:

• Not great for a precise definition of certain behaviors

There are two types of interaction diagrams typically used to capture the various aspects of interaction in a system:

• Sequence diagrams. A sequence diagram shows the interaction between objects in the sequence in which they take place. Sequence diagrams describe how the objects function within a system, and in what order, and are often used to document and understand what is required for new and existing systems.

• Collaboration diagram. Collaboration diagrams are used to define and clarify the roles of the objects that carry out a certain flow of events in a visual format, and serve as the main source of information when determining class responsibilities and interfaces.

7. Gantt charts

Gantt charts are a popular business process model for companies preparing for projects with distinct timelines, or that have time-sensitive processes that need to be captured and tracked. A Gantt chart plots a vertical list of tasks along a series of aligned, horizontal timelines. Each bar along the timeline represents the anticipated start and end date for a tasks, as well as its current status or progress.¹²

Strengths: 

• Quick overview of who is responsible for each task and when
• Easy for people involved in different parts of a process to see when they are meant to start and finish work
• Visualizes the whole project at once

Limitations:

• Includes little to no details on project scope and finances
• Does not visualize interactions between tasks or the overall shape of a process flow

8. Integrated definition for function modeling

Integrated definition (IDEF) for function modelling displays when parent activities give rise to child diagrams. There are 15 forms of IDEF and each addresses a different type of flow for functions, information, data, simulation model design, process description capture, and so on.¹³

Strengths: 

• Easy to understand, even without technical knowledge
• Useful for defining the scope of analysis

Limitations:

• Framework is very rigid to each use case

9. Colored Petri nets

When a system has numerous processes that interact and synchronise with each other, then colored Petri nets are ideal. This modeling technique is used to design, specify, simulate, and verify systems.¹⁴

Petri nets are unique in that they can represent both a state, such as passive, unsent, or waiting, and an action, such as send, receive, or transmit, in the same diagram. Colored nets use colors to differentiate their symbols, and use a formal, mathematical representation with well-defined syntax and semantics.

Strengths: 

• Mathematical foundations make it suitable for rigorous analysis
• Works well for concurrent activities

Limitations:

• Complexity can be difficult for beginners

10. Object-oriented methods

The object-oriented method of business process modeling is more than just modeling with objects: it encompasses message-passing, encapsulation (where internal detail is hidden), inheritance from class to subclass, and polymorphism (where the same procedure can operate on different data types).¹⁵

Strengths: 

• Modular structure promotes ease-of-use
• Features like inheritance make reusability easier

Limitations:

• More applicable for software development processes than general business processes

Think more strategically about your business process designs

Which business process modeling technique will you select for your business? Find the one that will ensure those involved in the system or process carry out their tasks in a consistent and efficient way, producing a predictable, measurable outcome.

Register now and join one of our project management courses or business management courses to learn more about the different types of process modeling available to your business.

MIT Sloan School of Management
Business Process Design for Strategic Management

University of Cape Town (UCT)
Foundations of Business Management online short course

Saïd Business School, University of Oxford
Oxford Leading Strategic Projects Programme

University of Cape Town (UCT)
Advanced Project Management online short course

The post 10 types of process modeling techniques explained, with examples appeared first on GetSmarter Blog.

“How do you regulate something that’s never existed?” This is the challenge governments face in its efforts to regulate the FinTech industry.  

In this video, Lauren Cohen and Christopher Malloy, Course Conveners in the Harvard VPAL FinTech online short course, discuss the battle between incumbents and start-ups, and consider which companies may be at a disadvantage when further industry regulation comes into effect.

With this in mind, they also reflect on who will lead the future growth of the FinTech industry – will start-ups continue to drive innovation in this sector, or will traditional financial institutions be at the forefront?

Transcript

How do you regulate something that’s never existed? That’s a classic question that technology industries have faced for many years, and now this war between government regulators, incumbents, and start-ups is forefront in the future of FinTech. What’s the right path for government regulation, and how should these FinTech start-ups and incumbents respond to this situation?

So, on this issue of government regulation, the question is, “When is the government going to crack down and when is it not?” And obviously in China, there’s been lots of crackdowns recently. Another case that’s been quite interesting here in the US, is this Madden versus Midland case. Where the government is coming in and potentially regulating FinTech lenders, and saying that they can’t charge above a certain interest rate.

The stated goal of that is, of course, to get rid of what’s called predatory lending, and help consumers from getting ripped off. And the question is, is FinTech ripping off these consumers or not? I mean you could make an argument that FinTech is coming in and helping the exact consumers that the government claims to be worried about.

Whenever you get into this government regulation issue too, then FinTech, are they going to be able to find a way around this? And if they can’t find a way around this, then do you have, Tech on FinTech, like FinTech squared, that’s going to come in, and be able to figure out a way to do it? However much the government wants to regulate these things, it seems like there’s always ways around these.

And so, it’s going to be tough for the government to regulate things like bilateral transactions. If you and I agree that: Okay, I’m going to lend you ten dollars at the beginning of the week, and you want to pay twenty at the end of the week, then of course that’s a massive interest rate. And yet, we might be happy to do that, right? One of us may really need the money until the other is happy to do that, and we’re both better off by making that transaction. So to the extent that the government comes in and tries to tamp down on these – could be not so great for their citizens.

Exactly, and one argument that FinTech companies will make is that these markets don’t exist, but they could exist at some price, and we’re making consumers better off, and – by creating this enormous market – society’s better off.

And so, then the question becomes, “Okay, who is most hurt by this regulation?”

Exactly. So are they going to come after the big established players? Or are they going to crack down the new incumbents?

If history has told us anything, it’s that these government regulators – they have jobs too, and so they want to prove that they’re really doing a great job, and it’s much more newsworthy to take down the big fish. It’s much more newsworthy, if you can take down that firm that everyone’s heard of. And so, in that sense, incumbents almost have more innovative risk in this space, than a new small innovator.

They have large established brands and client bases, and to the extent that they don’t want to put that at risk, they may be less willing to dive into some of these markets, and perhaps that’s why some of these FinTech companies have gotten market share in certain areas where the incumbents have been slow to enter, particularly in developed countries like the US.

But, then, you have this double-edged sword – if I’m late to enter into this market, there’s some kind of brand equity that can be built up from a LendingClub, or a Prosper, or one of these other people, then by the time I want to get into it, it’s too late.

There’s a big first-mover advantage. I can either try, and then compete late, or I try to partner with them. So this partnership angle is something that we’ve seen growing of late. The extent to which FinTech companies always want to build their own brand, and go public, and build a huge business, but now, realizing that maybe it’s in their best interest to partner, and of course, on the incumbent side, perhaps the same way. Like, building all this expertise in-house is quite costly and difficult. It might just be easier to establish a partnership.

You’ve seen this in other industries too, right? We’ve seen this in the biotech-farmer relationship, and how that’s evolved over time. And that, now, much of the innovation in that – in drugs, is been pushed down to biotech, and now large pharma companies – like a Merck or AstraZeneca – simply buy the biotechs later.

And have a marketing role.

Yeah, you can imagine something similar being true here. So, in that sense, what does government regulation do? And what do we need to worry most about in the FinTech space with government regulation? Well, it might not be that it’s going to massively impact the rate of innovation, and the technology that’s developed, but it impacts who does the innovation.

So the most valuable capital that you have, it turns out, is your human capital. So we want you to answer that big question: Where do you want to bet your career? Do you bet your career on that innovative new FinTech start-up that’s going to change the world, or with the incumbent firm that’s trying to get into this Fintech space?

The post How Regulation Works in FinTech appeared first on GetSmarter Blog.

Expensive infrastructure, employee training, and creating a culture of innovation are some of the challenges organizations face when implementing cybersecurity measures. In this video, Heather Adkins, Director of Information Security and Privacy at Google, discusses the critical role leaders play in driving cybersecurity awareness and developing risk mitigation strategies. 

The Harvard VPAL Cybersecurity: Managing Risk in the Information Age online short course equips you with the skills to protect the integrity of digital assets and lead your team through the complexities of risk management.

Transcript

Cybersecurity is going to be one of the biggest obstacles for organizations today and in the future.

The organizations that are succeeding at it are ones where their leaders are setting the tone and the cadence for the organization around cybersecurity. Especially a culture, where at every opportunity they’re taking the time to lay out how important it is to the organization.

I think the leader’s role in this is to educate themselves, of course, but also to inspire their organizations to focus on cybersecurity. It is a permanent function in their organizations, and only then will everyone feel like they can participate in that conversation. It is important that in security, we think of it not being just the role of experts, but the role of everyone. Everyone has a part to play.

C-suite leaders have a very particular role in the organization, with regard to cybersecurity. And I often tell my friends who are going into a CSO role, or a chief security officer role, to only pick the companies where you’re not going to have to push your agenda, but someone’s going to be asking for your agenda. Whether it be the board or whether it be the CEO. And to pick a role where you’re going to have that relationship and a bidirectional dialogue.

Cybersecurity is one of the hardest things the organization’s going to have to do. It’s expensive to continuously modernize your IT infrastructure, and it’s difficult to educate large workforces of people. And those things require a culture that supports change, and sometimes difficult change. And so, that relationship needs to be absolutely solid. And the CEO, the CIO, the CTO, and especially the CSO, need to all be on the same page about where you’re going.

If we are going to suppose that the C-suite leadership and middle management leadership own a part of the conversation and the dialogue, setting the tone and values for the culture, then I think that when you hire, you have to hire for your culture.

That means asking different kinds of questions. How do leaders, especially in the technology space, how do they see the role of technology and the human? I had a mentor who said, “For every new security thing we were going to put in place we were going to take away two.” And the idea there is to set a culture that says, “We’re going to be innovative about security, rather than traditional.” And, I think, it’s important that if you are trying to set that culture in your organization, that you have people thinking in these new and radical ways to make security a little bit easier for everyone in the company.

Make sure that your IT infrastructure is modernized, and that you’re keeping on top of that. We’re no longer in a time where we can buy something, and it’ll still be relevant ten years later. Technology turns over so quickly now and security improvements happen with each iteration.

Be modern in your fleet, as we call it, the machine fleet. And be modern in your thinking about how your employees interact with that technology. You’re going to need things like two-factor authentication and good cryptography. Let your roadmap really guide you, but don’t be afraid to push the boundaries of technology.

The post What Is a Leader’s Role in Cybersecurity? appeared first on GetSmarter Blog.

The overwhelming number of financial products on the market has given rise to two innovative FinTech firms. Wealthfront and OpenInvest provide a service that automatically selects the most suitable investment option based on the client’s preferences, goals, and social and ethical values.

In this video, Christopher Malloy, a Course Convener in the Harvard VPAL FinTech online short course, discusses how robo-advising has made it easier and more affordable for clients to invest and manage their portfolios.

But with a product that’s easy to replicate, how will these companies remain competitive in the long run? And how does this service impact the future of financial advisers?

Transcript

In this video, we will talk about the emergence of a new phenomenon in asset management, which is the idea of robo-advising. And we’ll focus in on two particular companies, Wealthfront and OpenInvest, who are pioneers in this space.

The simple reality of why this industry became into existence is because the wealth management industry is too complicated, it’s too opaque, and it’s too expensive. And those facts paved the way for the emergence of these firms. So now, on the expensive piece, a company called Vanguard, has been a pioneer in wealth management and making products that are very, very inexpensive. And so, Vanguard is now one of the largest wealth management firms in the entire world. But the problem with Vanguard is if you go to their website, you still have to decide which products to choose. And so, that search problem is actually quite hard, and I would encourage you to go onto their website and look at the thousands of products that they have on offer, and try to figure out what you would want to invest in.

And so, because of that problem, which we call the search problem, these robo-advising firms came into being. Wealthfront and OpenInvest basically grew right after the financial crisis, and the idea is quite simple, which is: Could you offer a product totally online, where someone could come onto the site and type in their risk preferences, things that they’re interested in, and goals, and could you then, as a firm, design a very automated product that would go and invest in cheap securities, like exchange-traded funds, and then automatically rebalance that portfolio such that the investor wouldn’t have to keep going and logging in, and they would in one click basically have their entire investment policy set by an automated firm?

So Wealthfront took that idea and designed a very appealing interface, targeted some younger investors, and grew dramatically since 2014, and became a big player in this robo-advising space. And then, seeing the success that Wealthfront had had, other firms began to enter. And one such firm was OpenInvest, which had a refinement of the idea of robo-advising based on this notion that, perhaps, investors had more preferences other than just risk and timeline, but maybe they cared something about, the environment, or maybe they cared about gun control, maybe they cared about some other social issues. So, they design an interface where people could screen out particular types of companies and particular types of investment, and that’s sort of capitalizing on this, the idea of socially responsible investing.

Now, the challenge of socially responsible investing, of course, is how do you define it? Because, for example, say you wanted to invest in environmentally friendly companies, well, then of course, naturally you’d think you would screen out Exxon and BP and these large oil stocks. And the problem with that is those are the very companies that are spending the most amount of resources and doing the most amount of R&D in the area of sustainable energy.

And so, the problem of how to measure social preferences and how to measure firms and investments that are socially responsible is actually quite difficult. And if there’s one thing we know from finance is that if you take a universe and you constrain it and make it smaller, and try to then pick the best investments within that, you are ultimately going to be sacrificing returns. And so, that’s this trade-off that OpenInvest is always having to think about, which is: Can I offer a product that delivers good returns while still screening out the things that people might not want to invest in?

Now, the other big challenges that both OpenInvest and Wealthfront face going forward is, because their products are so easy and transparent, they’re actually quite easy to replicate. You put in a couple parameters, and then they put you into some exchange-traded funds, and then there’s a machine that will automatically invest and automatically rebalance. And you can imagine this would be quite easy to replicate by larger, big financial firms, and so that’s an issue that they’re quite worried about.

A second issue is if we get to the point where all investment is done in a robotic way, what happens to all the financial advisers who are out there, and is it a good thing to get rid of all these financial advisers? So you might say it is, because they’re quite expensive, but at the same time, if you have machines trading with machines by machines, there’s almost no one at the end of the day who’s looking at the actual stocks, and pricing them, and doing the correct research to make sure that the prices are right. And so, you worry about a scenario where all humans are removed from the investment process. Will the prices be right, and what will happen to the returns on your investments going forward?

The post Approaches to Automated Investment appeared first on GetSmarter Blog.

Hackers, foreign influence, and ransomware are considered the most dominant sources of cyberattacks. However, disgruntled, disillusioned, or negligent employees may pose an even greater threat to a company’s data and networks. In this video, Debora Plunkett, former Director of Information Assurance at the National Security Agency, explains what actions organizations can take to mitigate the risk of insider threats. 

The Harvard VPAL Cybersecurity: Managing Risk in the Information Age online short course equips you with the skills to assess the threat posed by those closest to your organization.

Transcript

Insiders pose a threat to cybersecurity, first of all, because they have legitimate access to information, data, networks, storage. And, because they have that legitimate access, if they should choose to use it in an illegitimate way, they pose a huge risk. Because they could put company data at risk. They could put the company itself, the organization itself at risk. They understand the networks. They understand, in many cases, how data is stored. In many cases, they hold the keys to the kingdom.

If you think about the premise for cybersecurity, it is protecting data and information and the networks or infrastructures on which they reside. Insiders, because they have that legitimate access already, they cut the legs off a legitimate cybersecurity effort because they can go undetected, because they can move within an organization and not necessarily be noticed unless they take steps or perform actions that are unusual, and unless the organization’s actually looking for that.

Insiders target an organization’s information systems for a couple of reasons. Sometimes they are just disgruntled. They’re unhappy with the workplace, and they intend to perhaps leave and want to leave by doing harm to an organization. Sometimes they’re disillusioned. Socially disillusioned, politically disillusioned, and they might believe that the organization has information that can support their position or refute their position, one way or the other. And so, they would target information in an organization to embarrass an organization, to expose the inner workings of the organization.

Organizations can improve the acts of insiders, whether intentional or not, in a couple of ways. The first is that they have to understand what normal is and they have to be able to look for the abnormal. So, understanding what normal behavior is of an employee who has legitimate access and then recognizing when that individual takes an action that’s not normal – downloading large volumes of information or bringing equipment into the workplace that they might not otherwise have or use. Or working very late hours or odd hours are indicators that an employer should take a hard look at to see if there’s anything untoward happening.

The post How Insiders Can Threaten Cybersecurity appeared first on GetSmarter Blog.

Businesses, financial institutions, the government, and the public sector have reaped the rewards of advances in information technology. However, data is a valuable commodity and managing the risk of exploitation has become an industry-wide priority.

Transcript

Information technology has made huge advancements on the way that an organization can run.

However, with all of those advances comes new risks.

So, there are three important risks that I want you all to understand. The first is Business Operational risk, the second is Legal and Litigation risk, and the third is Reputational risk.

First, when you think about Business Operational risk, this literally means the risk that a cyberattack would take down your core operation in a way that prevented you from either delivering goods and services, accomplishing your mission, or – when it comes down to it in the private sector – just making money.

If you, for example, were targeted and you’re hacked, and you were no longer able to process payments, because the credit card system had been hacked, that has a big business impact.

You’re not able to process payments and, thus, you’re not able to bring in money and conduct sales.

Think about the Department of Defense, where I used to work, business operations are very important. We placed a high priority on our cybersecurity because we did not want to be in the situation where our network was down and, thus, we couldn’t fly aircraft; we couldn’t deliver precision munitions.

Also, think about it in the banking sector. And a good case is the case of the Iranian attack on the financial services, where they conducted DDoS, which prevented customers from coming to their public-facing website, process payments, and get their business done.

The next thing I’d like to talk to you about is Litigation risk and Legal risk. So, when one of these cyberattacks happens, really, one of the things that can pose the biggest risk to a private sector firm, in particular, is Litigation risk. That means that some of the activities that resulted from the hack will result in lawsuits against your firm and your executives.

In Target, for example, there was a shareholder derivative suit. That meant all the holders of Target stock sued the board and the executives of Target because they felt that their reaction to the attack was not good. That costs the firm money and then lowers their profit.

Another form of Legal risk is Criminal Legal risk. When there’s a hack, usually, some aspects of the criminal code is influenced and implicated. And that means you’re going to be dealing with the FBI, or other law enforcement agencies, depending on how it’s set up in your country.

That also can present a reputational risk problem because now people see that you’re mixed up with a criminal lawsuit.

Finally, I want to talk to you about Reputational risk. This is really important to understand. And it really drives and flows from those first two risks that I talked to you about.

So, for example, if you’re hacked and your business operations go down, then you can expect that your reputation is going to suffer because customers are accustomed to being able to depend on you for the service.

In the case of the financial services who were attacked by the Iranians; consumers who were used to be able to go their webpage, do all their banking transactions and then move on. When they can’t do that, their opinion of you and your firm goes down, it diminishes, they start to question whether you’re a credible organization.

Here’s another perfect example that can have a real financial impact: Yahoo disclosed major hacks over a period of time that gave the impression to people who were considering doing a merger and acquisition with Yahoo that they weren’t competent in their handling, either of their cybersecurity, the litigation, or the public affairs. It hurt their reputation as a firm to the point that it really drove down the cost and the transaction.

So, reputational risk is something that seems intangible, but is really important, and something that you need to think about consciously.

The best way to really understand these risks that I’ve just talked about is for you to think critically about your own work now, maybe your own firm, if you work in the government, your organization’s mission. And think through, from your perspective: How could these risks come into play with where I work right now?

The post Risks and Considerations of Cybersecurity appeared first on GetSmarter Blog.

Before good and useful ideas can be brought to market, companies need a way to generate novel and innovative ideas. Toyota’s chief innovations officer decided to create a company-wide innovation fair, inviting all employees to share their ideas for innovations that could be important to Toyota. Over the years, this innovation fair developed into many different models whereby new ideas could be explored and funded within the organization. Like Toyota, your company can be transformed through business design and strategy inspired by technology. 

Transcript

The real challenge is learning whether customers will actually pay for your good ideas. But you might be asking, “Wait a minute, where do these good ideas come from?” So let’s take a look at how companies have developed a set of good ideas of things that technology could do for them.

Let me start with the example of Toyota. Years ago, the CIO said, “You know, we’re good at incremental innovation, but I feel like we have innovative thinkers all over this company, and we’re not taking advantage of their ideas.”

So he encouraged everyone in his organization to imagine an innovation that could be important to Toyota. They could show anything they wanted; it would be kind of like a science fair. They would exhibit this on Innovation Fair Day. Now to generate excitement in the company, he called it a big event, the Technology Innovation Fair – everybody come. And to make sure people really showed up, he promised ice cream at the end of the Innovation Fair. 

What happened is a number of people found this a very exciting idea and did indeed display innovative ideas, and even more people thought it would be fun to go see what the possibilities were. And, in fact, they had so many people show up that as they all got to the ice cream room, it got way too crowded and hot, all the ice cream melted and that part of the Innovation Fair was, in fact, not a success. But the Innovation Fair itself was a rousing success. In fact, not only did they have some ideas that people with money wanted to fund – I’m talking about people within Toyota wanted to fund, and indeed they started to fund experiments that would test ideas – but, on top of that, people who are not in IT started saying, “Wait a minute. What about us? We have innovative ideas too.” So the Innovation Fair concept started to grow; it was an annual event, not just for IT, but for the entire organization. 

What grew out of that was more and more ideas that seemed worth testing. Initially, they only funded the winners. But then they realized there were actually a lot of good ideas out there. So they started to invest in more ways to identify and test ideas. They created a Shark Tank idea. A Shark Tank being the television program where innovators or entrepreneurs get up and say, “I have this really cool idea. Do you want to fund it?” And there would be people with budget, in Toyota, listening to these ideas, challenging them, and in some cases funding them. They also created a little venture fund, where the CIO could identify some good ideas worth testing. 

Understand, you do not want to start where Toyota is today, with multiple innovation labs doing many experiments and, as many as 600 people. You want to start like Toyota started, with an innovation fair that grows into more and more sharing of ideas, testing, experimenting and then you too will grow into a bigger model – perhaps like Toyota’s, perhaps quite different. 

The idea is to learn how to generate ideas, enrich those ideas, and test them with customers so you learn what it is they love. 

The post Toyota’s Approach to Innovation appeared first on GetSmarter Blog.