Business Analyst Community & Resources | Modern Analyst

What are agentic workflows?

Adrian M. — Mon, 29 Dec 2025 21:07:00 GMT

An agentic workflow is a business process where an AI system doesn’t just suggest work—it can plan, decide, and take actions to move a task toward completion.

In a traditional workflow, software follows explicit rules and humans perform decisions. In an agentic workflow, an AI “agent” can:

interpret intent (goal),

break work into steps,

choose tools/actions,

execute actions (sometimes across multiple systems),

and escalate to humans when needed.

Why it matters to Business Analysts

Agentic workflows change what “requirements” must cover. It’s no longer enough to define screens and rules—you must define:

Decision rights: what the agent is allowed to do without approval

Boundaries: what it must never do (data, actions, users, systems)

Escalation: when it must stop and hand off to a human

Evidence: logs, explanations, and auditability (what happened and why)

Example

A “copilot” might draft a customer response for a human to send.
An agentic workflow might:

read the ticket,
pull order history,
decide if a refund is warranted,
issue the refund in the payment system,
notify the customer,
and log the full decision trail.

What a strong BA delivers

A workflow map showing human + agent handoffs

Risk tiers (low/medium/high) tied to required controls

Acceptance criteria for audit logs, monitoring, and rollback (“kill switch”)

Give an example of using process-mining insights to justify a re-engineering project

Adrian M. — Tue, 08 Jul 2025 04:22:00 GMT

Scenario
A mid-size mortgage lender complained that “simple” refinance loans were still taking 22 days on average—well above its advertised 15-day target. Management suspected underwriting staffing was the bottleneck and planned to add headcount. Before approving the budget, the analytics team ran a process-mining study on six months of event logs exported from the loan-origination system (LOS), e-signature platform, and credit bureau calls (2.4 million time-stamped events).

1. Discovering the Real Process

Process-mining software reconstructed the actual flows and overlaid performance metrics:

Median lead time: 22 days but only 7 hours of value-adding work.

Variant explosion: 138 distinct paths; the top 3 covered just 46 % of cases, signalling uncontrolled exceptions.

Rework loops: 38 % of loans cycled through “Document Requested → Upload → Missing Docs” at least twice, adding 3.7 days per loop.

Idle gaps: After “Credit Pull,” 62 % of files waited >48 hours for an initial review—even for low-risk FICO ≥ 740 borrowers.

2. Root-Cause Insights

Drilling into conformance checks showed that “Missing Docs” loops were highly correlated with loans sourced by two particular call-center teams. Text-mined notes revealed those agents were using an outdated checklist. Idle gaps traced to a rules engine that routed every file—even low-risk ones—to the same underwriting queue.

3. Re-engineering Proposal

The BA presented a two-pronged redesign:

Straight-Through Processing (STP): Auto-approve credit pulls ≥740 FICO & <80 % LTV for an “expedited” queue; introduce robotic validations for income and asset docs.

Checklist Modernization: Deploy an in-app wizard that enforces the current documentation checklist before a file can be submitted.

Projected benefits were simulated in the process-mining tool by removing rework loops and idle events:

Metric	As-Is	Simulated Future State
Avg. Lead Time	22 days	11 days
Rework Rate	38 %	8 %
Underwriter Touches	2.6	1.1

The model showed a 33 % capacity release—worth more than three new underwriters.

4. Business Case & Sign-off

By quantifying each redesign lever with log-level evidence, the BA framed a clear ROI: $1.2 M annual savings (labor + lock-extension fees) vs. $300 k one-time platform changes. The executive team approved the re-engineering project, cancelling the headcount request and setting a KPI to monitor the live process via continuous process mining.

Take-away

Process-mining moved the conversation from anecdotal “we’re swamped” to data-driven proof, allowing the BA to justify a targeted, automation-first re-engineering effort that halved cycle time without adding staff.

What is Process Mining?

Adrian M. — Tue, 01 Jul 2025 04:13:00 GMT

Process Mining is a data-driven discipline that extracts digital footprints from the event logs of enterprise systems (ERP, CRM, BPM, ticketing, IoT, etc.) and automatically reconstructs how work actually flows through an organization. Unlike workshop-based mapping that relies on interviews and sticky notes, it reveals the real sequences, variants, wait times, and exceptions recorded by time-stamped transactions.

Process Mining turns system exhaust into x-ray vision for business processes, empowering analysts to diagnose, quantify, and continuously manage improvement opportunities with scientific precision.

Three Core Capabilities

Discovery – builds a visual as-is model from raw logs, highlighting the most common path and every deviation.
Conformance – compares that empirical model to a reference design or policy to spot violations, rework loops, and compliance risk.
Enhancement – enriches the model with performance data (cycle time, cost, carbon, customer sentiment) so improvement scenarios can be simulated.

How It Works: Ingest logs → identify case ID, activity, timestamp → algorithmically create process graph → overlay metrics → drill down by variant, root cause, or persona → export insights to BI or automation tools.

Common Metrics Reported: Lead and touch time, waiting time, throughput, first-pass yield, variation count, compliance score, automation potential.

Tools: Commercial platforms (Celonis, UiPath Process Mining, Apromore, Minit) and open-source libraries (ProM, PM4Py) ingest logs from SAP, ServiceNow, Salesforce, or custom databases via connectors or SQL.

Why Business / Systems Analysts Use It

Cuts through hearsay: replaces “I think we do X” with provable facts.

Quantifies waste: idle time, rework, ping-pong hand-offs, and variant sprawl are measured in minutes and dollars.

Builds bulletproof business cases: simulation shows how removing one approval or adding a bot affects lead time and capacity.

Feeds requirements: the mined “happy path” and exceptions translate directly into user stories, SLAs, and test scenarios.

Supports continual monitoring: dashboards alert stakeholders when the live process drifts from the improved design.

Example

A telecom analyzed 12 million trouble-ticket events. Mining revealed that 27 % of tickets detoured back to Level-1 support after Level-2 touched them, adding 2.1 days per case. Automating ticket triage and adding a knowledge-base prompt cut rework to 4 % and saved £3 M annually—numbers the analyst could present with confidence because they came straight from the logs.

How do you decide whether to optimize or re-engineer a process?

Adrian M. — Sat, 10 May 2025 01:26:00 GMT

When I’m deciding whether a process just needs a tune-up or a complete tear-down, I start with a quick reality check: Is the process basically sound but sluggish, or is it fundamentally getting in the way of our goals?

First, I look at performance metrics and complaints. If we’re missing targets by, say, 10–15 percent and the issues cluster around clear bottlenecks—extra approvals, manual re-keying—there’s usually mileage in streamlining what we already have. Small fixes like removing duplicative steps or adding a bit of automation can get us back on track quickly and at low risk.

If the gap is big—think 30 percent or more—or if the pain points come from deeper problems such as outdated tech, new regulations, or a process that just doesn’t fit how the business operates now, patching won’t cut it. In that case, I lean toward re-engineering: stepping back, redefining the objectives, and designing a fresh workflow that’s future-proof.

Two other things help me choose. Cost and risk: I run a rough cost-versus-benefit comparison. If incremental tweaks deliver most of the value for a fraction of the effort, optimization wins. But if maintaining the old process exposes us to fines, lost customers, or tech debt, a reboot makes more sense. Culture matters too. In a team comfortable with steady, Kaizen-style improvements, optimization sticks. If leadership is pushing a wider transformation, people often have the appetite—and the air cover—for radical change.

So, in short, I optimize when targeted tweaks will meet the goal, and I re-engineer when the current architecture blocks the business from moving forward. It’s always a balance of impact, cost, risk, and readiness.

What is Customer Journey Mapping?

Adrian M. — Sun, 29 Dec 2024 23:58:00 GMT

Customer Journey Mapping is a strategic tool used by businesses to visualize and understand the process that customers go through when interacting with a company, from initial awareness to post-purchase experience. It provides a detailed, visual representation of the steps, emotions, touchpoints, and actions a customer takes during their relationship with a brand. This map helps businesses gain insights into the customer experience and identify opportunities for improvement, optimization, and innovation.

The primary goal of customer journey mapping is to align a company’s efforts with the needs, pain points, and expectations of customers. By analyzing the full spectrum of a customer’s experience, from discovery to loyalty, businesses can make more informed decisions about how to enhance their products, services, and customer interactions.

A typical customer journey map includes several key components:

Customer Personas: These are detailed descriptions of the target audience segments, including demographics, behaviors, and motivations.

Stages of the Journey: The journey is divided into distinct stages that reflect a customer’s path—such as awareness, consideration, purchase, retention, and advocacy. Each stage represents different touchpoints where the customer engages with the business.

Customer Touchpoints: These are the specific points of interaction between the customer and the company, which can include websites, social media, emails, customer service, or in-store experiences.

Customer Emotions: Understanding how customers feel at each touchpoint helps identify pain points and areas for improvement. A positive emotional experience can build loyalty, while frustration or confusion may lead to churn.

Opportunities and Gaps: Mapping allows businesses to identify areas where customers face difficulties or where there are gaps in the journey that could be improved or enhanced, ensuring a more seamless and enjoyable experience.

By employing customer journey mapping, businesses can better anticipate customer needs, personalize interactions, and create a more cohesive experience across all touchpoints. Ultimately, this leads to higher customer satisfaction, increased loyalty, and improved business performance.

In a business analyst context, customer journey mapping is essential for making data-driven decisions, improving customer service, and aligning cross-functional teams to meet customer expectations effectively.

Example of Customer Journey Map

Attribution: Wikimedia.org under CC BY-SA 4.0

What is Composable Business Architecture?

Adrian M. — Sun, 10 Sep 2023 22:50:00 GMT

Composable Business Architecture (CBA) is a business architecture approach and methodology that emphasizes flexibility, adaptability, and modularity in designing and organizing the components of a business. It is a response to the ever-changing and complex nature of the business environment, where organizations need to be agile and responsive to rapidly evolving market conditions and customer demands.

“Composable business means creating an organization made from interchangeable building blocks” - Kasey Panetta, Gartner Contributor

The same way a Lego brick is the basic building block of any Lego masterpiece, a Packaged Business Capability (PBC) is the basic building block of a composable business model. Each of the enterprise PBCs fulfills a critical function of the business.

“Composable business is a natural acceleration of the digital business that you live in every day. It allows us to deliver the resilience and agility that these interesting times demand.” - Daryl Plummer, Distinguished VP Analyst, Gartner.

Gartner identifies, four principles of composable business architecture:

Modularity - dividing the business domain into smaller discrete components with clearly defined responsibilities

Autonomy - ensure the individual components are as independent as possible from other components in order to minimize dependencies and enterprise impact when one component needs to be changes or replaced

Orchestration - the process and tools used to control the flow and collaboration among the individual business components in order to achieve specific business objectives

Discovery - the ability to easily find and understand existing business process and component assets within the enterprise to facilitate reusability and for quickly composing new processes for in response to opportunities or business landscape changes

“An organization that delivers business outcomes and adapts to the pace of business change. It does this through the assembly and combination of packaged business capabilities… application building blocks that have been purchased or developed.” — Gartner

When it comes to establishing composable businesses, Gartner defines three essential building blocks, as follows:

Composable thinking - Composable thinking is an adaptable problem-solving approach that involves breaking complex challenges into modular components, allowing for flexible and creative recombination to address evolving needs, fostering innovation and efficiency. It emphasizes modularity, reusability, and customer-centricity in finding solutions and is relevant across diverse domains, including technology, business, and product development.

Composable technology - Composable technology refers to an IT infrastructure and architecture approach that leverages modular components and software-defined resources, enabling organizations to dynamically and flexibly allocate computing, storage, and networking resources to meet specific application or business requirements. This approach enhances agility, scalability, and resource utilization in a rapidly changing technology landscape, allowing organizations to adapt quickly to evolving needs and opportunities.

Composable business architecture - Composable Business Architecture is a strategic framework for designing and organizing a business's components, such as processes, capabilities, and technologies, in a modular and flexible manner. It enables organizations to quickly adapt to changing market conditions, customer demands, and emerging opportunities by promoting modularity, reusability, and interoperability in their business operations.

Composable Business Architecture is particularly relevant in industries characterized by rapid change, digital transformation, and the need for continuous innovation. It helps organizations stay competitive and responsive to emerging opportunities and challenges by providing a framework for designing and organizing their business components in a flexible and adaptable manner.

What is BPMN?

Chris Adams — Fri, 02 Jun 2023 01:37:00 GMT

BPMN stands for Business Process Model and Notation, it is an industry standard that provides businesses with the capability of visualizing and communicating their internal business processes and their external business to business processes in a standard manner. Beyond the obvious use of modeling business processes, BPMN has been created with the key goal of creating a bridge between the business process modeling notation (BPMN) and IT-oriented execution languages that will implement the modeled business processes within a business process management system.

This is done by maintaining and supporting a strict internal model that maps the rich set of graphical objects and object attributes of BPMN to executable languages such as the Business Process Execution Language (BPEL) or Business Process Modeling Language (BPML) to support immediate execution of modeled buiness processes.

Key features and concepts of BPMN include:

Symbols and Notation - BPMN utilizes a set of standardized symbols and graphical elements to represent various elements of a business process, such as tasks, events, gateways, and flows. These symbols are intuitive and can be easily understood by both technical and non-technical stakeholders.
Process Modeling - BPMN allows for the modeling of end-to-end business processes, capturing their flow, activities, decisions, and interactions. It enables the representation of complex processes, including parallel and sequential flows, loops, and conditional paths.
Collaboration and Choreography - BPMN supports the modeling of collaborative processes involving multiple participants or organizations. It allows for the representation of interactions, message flows, and responsibilities between different process participants in the BPMN collaboration diagram.
Event-driven Modeling - BPMN includes specific symbols and constructs to represent events that trigger or result from certain activities or states in a process. It enables the modeling of event-driven behavior, such as timers, message events, and error events.
Process Governance and Analysis - BPMN provides a foundation for process governance, documentation, and analysis. It enables stakeholders to document and communicate process requirements, analyze process performance, identify bottlenecks, and optimize process flows.

By leveraging BPMN, business and systems analysts can effectively analyze, communicate, and optimize business processes, collaborate with stakeholders, facilitate requirements elicitation, simulate process behavior, support system integration, and ensure compliance with regulations.

What is Human In the Loop (HITL)?

Adrian M. — Sat, 11 Mar 2023 21:45:00 GMT

Human in the loop (HITL) refers to a design or operation process or system where human intervention or oversight is involved in certain activities that may also utilize machine learning, artificial intelligence, or automation. In other words, HITL is a concept that describes the collaboration between human decision-making and machine intelligence.

In “human-in-the-middle” systems, humans are responsible for supervising, monitoring, and directing the actions of machines or software algorithms. Human oversight is critical for ensuring that the automated systems are working correctly, making ethical decisions, ensuring the legality of the action, and achieving the desired business outcomes. The human in the loop concept can be used in a variety of applications such as back office process systems, autonomous vehicles, medical diagnosis, fraud detection, and many more.

The HITL approach is often used in situations where it is crucial to balance the benefits of automation with the limitations of machines. By incorporating human oversight, "human in the loop" can enhance the accuracy, efficiency, and safety of automated systems while reducing errors and risks.

Some quick points about the human in the loop approach:

HITL is a model of human-machine interaction where a human is involved in the decision-making process of a partially automated system.

In HITL systems, a human user oversees and monitors the automated process and makes final decisions based on the results produced by the system.

HITL is commonly found in fields such as machine learning and artificial intelligence (AI).

HITL helps ensure that the system is making accurate and appropriate decisions.

HITL is a compromise between fully automated systems and fully human-driven systems.

HITL improves the speed and accuracy of decision-making while also ensuring that important factors such as ethics, morality, and context are taken into account by the human.

How might a business analyst use BPMN differently for Business Models than for Executable Models?

Chris Adams — Sun, 10 Apr 2022 16:20:00 GMT

The origins of BPMN began in the area of executable models. That is, it was created to be precisely interpreted by workflow engines or business process management systems in order to automatically orchestrate how information, documents, or other workflow items are directed through a system. The benefit of an executable model is that it can be changed and immediately re-executed to establish a new workflow. At least, that’s the idea.

Why bother with identifying use cases? Why not jump straight to process flows?

Chris Adams — Thu, 07 Oct 2021 16:15:00 GMT

Whether documenting business processes or systems processes, use cases and process flows can both be valuable tools.

A use case description textually describes the main success scenarios of a process as well as the major alternate scenarios and exception scenarios. Process flows do the same thing, though in graphical form. A process flow documents the step by step activities of the main success scenario. Decision points (often shown as diamond shapes) show where the main scenario branches off into alternate scenarios.

What is JTBD - Jobs to be Done Framework?

Adrian M. — Sat, 11 Sep 2021 06:37:00 GMT

“Jobs to be Done” also known as JTBD is a market and customer research framework with the goal of identifying the underlying tasks (aka jobs) that customers are trying to get done in the given problem domain.

The Jobs to be Done theory and method was first popularized by Clayton Christensen, Harvard Business School professor, in his very popular book The Innovator’s Solution.

In one of his article titled “Know your Customer’s Jobs to be Done”, Mr. Christensen clearly explains the genesis and core of this concept:

“The theory of jobs to be done was developed in part as a complement to the theory of disruptive innovation—which at its core is about competitive responses to innovation: It explains and predicts the behavior of companies in danger of being disrupted and helps them understand which new entrants pose the greatest threats.”

“What they really need to home in on is the progress that the customer is trying to make in a given circumstance—what the customer hopes to accomplish. This is what we’ve come to call the job to be done.”

To even begin to understand your customer you need to walk a mile in their shoes; you need to understand the jobs they have to do aka their “jobs to be done”. Tasks or jobs are foundational for discovering what motivates your users/customers and why they behave the way they do.

New Markets Advisors, a leading consulting firm in this area, uses a proprietary step by step approach to their Job Roadmap when they apply the Jobs to be Done method to help their customers to innovate:

Establish Objectives -> Strategy

Plan Your Approach -> Research Methods

Build a Jobs Atlas -> Customer Research

Generate Ideas -> Ideation

Reframe Your Perspective -> Open Innovation

Experiment & Iterate -> Test and Learn

Other practitioners take a more traditional approach to jobs-to-be-done by thinking of job as a process which is documented using a Job Map. Basic steps to create a Job Map include:

Define - Define the aspects of the job your customer has a problem with.

Locate - Locate and define the inputs and resources your customer needs to do the job.

Prepare - Identify how your customer must prepare and transform the inputs in order to get the job done.

Confirm - Ask what your customer needs to verify or validate to ensure the job is done right.

Execute - What does the customer need to do to execute the job successfully.

Monitor - Determine what areas or metric your customer needs to monitor to ensure the job is done right.

Modify - What aspects of the job steps need to be changed in order to successfully complete the job.

Conclude - Are there any specific steps which must be completed in order for the job to be final and done?

Business analysts and product managers can use the Jobs-to-be-Done (JTBD) method to create innovative product roadmaps.

How would you build a Business Process Model?

Chris Adams — Mon, 09 Nov 2020 03:31:00 GMT

Building a Business Process Model is a complex task that requires a number of successful steps to complete.

1) Determine scope

2) Gather background information

3) Conduct interviews

4) Begin Modeling

5) Validate and iterate

What is DMN and how is it used to support BPMN?

Chris Adams — Sun, 11 Oct 2020 17:35:00 GMT

BPMN is used to define business processes as a sequence of activities. Gateways are used to show branching of different process paths. For many years, analysts would clumsily model decision logic directly in business process models in an attempt to fully define process branching logic. This made process models messy.

DMN or Decision Modeling Notation was published in 2015 by the Object Management Group. It's a graphical language for specifying business decisions. DMNs primary purpose is to give analysts a tool for separating the business decision logic from the business process.

Can you have multiple start and end events in BPMN?

Chris Adams — Fri, 27 Jul 2018 14:40:00 GMT

The BPMN standard allows for multiple start and end events to be used at the same process level. There a few reasons why the business analyst may use multiple start or end events, but caution must be used to avoid mistakes.

What are the 5 basic categories of elements in BPMN?

Chris Adams — Mon, 13 Mar 2017 15:25:00 GMT

BPMN is a robust notation designed to balance two competing needs. The notation should be simple enough for all stakeholders to understand, yet robust enough to handle complex orchestration of events to a level of detail which can be made executable. Not an easy thing to do. However, by organizing elements into distinct categories, a sizable notation can be more easily understood.

What is the difference between a task and an activity in BPMN?

Chris Adams — Sun, 08 Sep 2013 22:39:00 GMT

Just like all squares are rectangles but not all rectangles are squares, similarly all tasks are activities but not all activities are tasks.

In BPMN terminology, activities represent work completed by a group or organization. A process, sub-process, or task are all types of activities. Some activities are atomic (a task) while others are not atomic (process and sub-process) since they can be further decomposed.

So you can use the generic term activity when discussing processes and work being done. But if you want to be less ambiguous in your meaning the terms process, sub-process, and task should be used.

Describe the BPMN Complex Gateway and how it is used in process modeling?

Chris Adams — Sat, 06 Jul 2013 20:47:00 GMT

BPMN Complex Gateways can be used for splitting or merging a complicated process flow scenario. Complex Gateways are not often used in BPMN since the simpler and more well understood parallel, inclusive, and exclusive gateways handle most situations. For a refresher of how each of these gateways work you can review the following:

BPMN Parallel Gateway

BPMN Inclusive Gateway

BPMN Exclusive Gateway

However, for more involved scenarios that use complex sequence flow rules the BPMN Complex Gateway can actually simplify your model by eliminating complicated combinations of multiple gateways.

A Complex Gateway allows the modeler to specify any arbitrary rules needed to determine how to emit tokens or process arriving tokens. The example below shows how the BPMN Complex Gateway could be used for an interview process where references have been requested for a particular candidate. The process is intended to request 5 references, but for the sake of time it will proceed after receiving only 3.

This example diagram shows the process with a placeholder (dotted diamond) for a gateway.

If no gateway is used then the flow is never synchronized and the tokens will continue on to the next task as each token is received. This is not how we want the process to work. An exclusive gateway would behave the same way when used to merge flows. If a parallel gateway is used then the gateway would wait for all 5 tokens to arrive before proceeding. This also is not how we intend for the process to work. Similarly, an inclusive gateway waits for all tokens that were emitted from the corresponding splitting gateway. Therefore, in this instance the best solution is the Complex Gateway.

The complex gateway is coupled with an annotation where the specific rules for processing the tokens are described. We can clearly state that the process will proceed once 3 (of the possible 5) references are received.

Bigazi provides a nice YouTube simulation of the Complex Gateway

Describe the BPMN Event-Based Exclusive Gateway and how it is used in process modeling?

Chris Adams — Sat, 06 Jul 2013 20:34:00 GMT

The more common BPMN gateways that you might be familiar with are usually data-based gateways such as the inclusive and exclusive gateways. Here is the list of all the BPMN Gateways.

When a data-based exclusive gateway is used, the value of a data element is evaluated by a condition statement to determine which path the process should proceed along.

In the following example, when the process reaches the gateway it already has data regarding the distance that must be traveling. The gateway condition evaluates the data to determine which path should be taken.

BPMN also provides event-based gateways. Event-based gateways do not evaluate data. Instead, when the Event-Based Exclusive Gateway is reached it waits. In particular it is waiting for an event which it will use to determine which path the process should proceed along. An event-based gateway is required any time a decision is made by another participant based on data that is not visible to our process.

If you examine the icon used for the BPMN Event-Based Exclusive Gateway you will see that it inherits elements from several other BPMN icons. First, there is the diamond which clearly shows it’s a gateway. Per the BPMN specification, this gateway is only used to branch the process and is not used for merging or synchronization. Within the diamond shape, there is a thin double circle representing an intermediate event shape. Finally, there is a pentagon shape which represent multiple events. So the Event-Based Exclusive Gateway is a gateway that deals with multiple intermediate events.

In the following example, the Event-Based Exclusive Gateway waits for an event to arrive, passes the flow of control along the correct path, and then doesn’t permit any other tokens to be passed along. Any other events that arrive as part of the same process instance will be ignored.

The event-based gateway also has the ability to be used to instantiate a process. When this is the case the Event-Based Exclusive Gateway icon has only a single circle within the diamond. When used to start a process, the Event-Based Exclusive Gateway allows the process to start in a number of ways based on the event that triggers it.

For more details on the Business Process Modeling Notation (BPMN) including BPMN Gateways, please reference our Cheat Sheet for BPMN (Business Process Modeling Notation).

Describe the BPMN Event-Based Parallel Gateway and how it is used in process modeling?

Chris Adams — Sat, 06 Jul 2013 20:23:00 GMT

The more common BPMN gateways that you might be familiar with are usually data-based gateways. These include the inclusive and exclusive gateways. Here is the list of all the BPMN Gateways.

When a data-based gateway is used, the value of a data element is evaluated by a condition statement to determine which path the process should proceed along.

BPMN also provides event-based gateways. Event-based gateways do not evaluate data. Instead, event-based gateways wait for an event which is then used to determine which path the process should proceed along. An event-based gateway is required anytime a decision is made by another participant based on data that is not visible to our process.

Sometimes, the start of a process is dependent upon several events having occurred. In this instance a Event-Based Parallel Gateway is used.

If you examine the icon used for the BPMN Event-Based Parallel Gateway you will see that it inherits elements from several other BPMN icons. First, there is the diamond which clearly shows it’s a gateway. Per the BPMN specification, this gateway is only used to branch the process and is not used for merging or synchronization. Within the diamond shape, there is a single circle representing a start event. Finally, there is a plus shape which represent that all of the following events should occur in parallel. So the Event-Based Parallel Gateway is a gateway that waits to receive multiple start events and allows them all to proceed in parallel.

It is important to note that while the Event-Based Parallel Gateway will allow multiple events to pass through and start the corresponding portion of the process, it does not wait for all of the events to arrive. That is, it does not wait and synchronize the events before the start of each process path is permitted.

For more details on the Business Process Modeling Notation (BPMN) including BPMN Gateways, please reference our Cheat Sheet for BPMN (Business Process Modeling Notation).

Describe the BPMN Exclusive Gateway and how it is used in process modeling.

Chris Adams — Sun, 12 May 2013 00:02:00 GMT

An Exclusive Gateway is one of the most common gateways used in BPMN and is depicted as an empty diamond or can be show as a diamond with an “X” within it. The empty diamond is most commonly used by modelers. As a best practice, once the modeler chooses which symbol to use he should be consistent throughout the diagram.

When used as a diverging gateway (splitting the sequence flow into many paths), an exclusive gateway can have 2 or more outgoing paths. It is used to direct the sequence flow along a single path where the condition evaluates to “True”. When using an exclusive gateway ONE AND ONLY ONE path can be taken for a given instance of a process which is why the term exclusive is used. Each condition is checked until a condition evaluates to true. Once a condition evaluates to true the flow proceeds along that path (a single token is emitted) and no other conditions are checked.

Often, there is one path which is identified as the default path. This means that if no other path condition evaluates to true then the default path will automatically proceed. If the process modeler fails to define a default path and none of the path conditions evaluate to true then a runtime exception occurs.

When used as a converging gateway, an exclusive gateway will pass along any token it receives without synchronization. This should intuitively make sense since the exclusive gateway is most commonly used to merge the paths which came from a diverging exclusive gateway where only one condition is permitted to evaluate to true and emit a token. If the converging exclusive gateway waited for other tokens for synchronization, it would be waiting forever.

Describe the BPMN Parallel Gateway and how it is used in process modeling.

Chris Adams — Sat, 11 May 2013 23:42:00 GMT

A Parallel Gateway is one of the most common gateways used in BPMN and is depicted as a diamond with a “+” within it.

When used as a diverging gateway (splitting the sequence flow into many paths), a parallel gateway can have 2 or more outgoing paths. When using a parallel gateway ALL paths must be taken for a given instance of a process. Therefore, there are no conditions checked before proceeding along each path.

When used as a converging gateway, a parallel gateway will wait for all tokens to arrive from each incoming path before merging and synchronizing the flow.

Describe the BPMN Inclusive Gateway and how it is used in process modeling.

Chris Adams — Sat, 11 May 2013 23:29:00 GMT

An Inclusive Gateway is one of the most common gateways used in BPMN and is depicted as a diamond with a circle within it.

When used as a diverging gateway (splitting the sequence flow into many paths), an inclusive gateway can have 2 or more outgoing paths. It is used to direct the sequence flow along all paths where the condition evaluates to “True”. When using an inclusive gateway ONE, SEVERAL, OR ALL paths can be taken for a given instance of a process. Each condition is checked and if it evaluates to true a token is sent down the path. With an inclusive gateway, even after a condition evaluates to “True” the rest of the conditions are checked to see if other tokens should be released along the other paths.

When used as a converging gateway, an inclusive gateway will wait for all tokens to arrive before merging and synchronizing the flow. Each converging inclusive gateway needs to have an associated diverging gateway earlier within the process such that it is aware of which paths were activated and the tokens it should be waiting for.

What is the difference between an interrupting event and non-interrupting event in BPMN?

Chris Adams — Fri, 05 Apr 2013 23:01:00 GMT

To understand the difference between a BPMN Interrupting Event and a BPMN Non-interrupting Event it is best to first understand the definition of a boundary event.

A boundary event is an event shown on an activity boundary that can be triggered or thrown at any time while the associated task or activity is being performed. Below, a Timer boundary event is shown. Some boundary events are time based while others are based on a condition which must resolve to ‘True’ in order for that branch of the flow to proceed.

When an activity is being performed and a boundary event is triggered or thrown, a token is sent along the path defined by the boundary event. However, the question remains as to what should happen with the existing activity that was being performed.

In the case of an Interrupting Event, the activity which was being performed will immediately be canceled. An Interrupting Event is shown by a solid circle or two solid circles around the event icon depending on whether the event was a start event (one solid circle) or an intermediate event (two solid circles).

In the case of a Non-Interrupting Event, the activity which was being performed will continue in parallel along with the new flow that was initiated by the boundary event. The current activity will NOT be cancelled (or interrupted). A Non-Interrupting Event is shown by a dashed circle or two dashed circles around the event icon depending on whether the event was a start event (one dashed circle) or an intermediate event (two dashed circles).

You can find the BPMN Event Notation here.

Describe the different Timer Events that are used in BPMN.

Chris Adams — Fri, 05 Apr 2013 22:41:00 GMT

There are 4 variations of Time Event symbols that can be used in BPMN. Before describing how each is used, a quick reminder of the overall BPMN Event Notation is helpful.

The 4 BPMN Timer Event types are:

Timer Start Event, Interrupting

Timer Start Event, Non-Interrupting

Timer Intermediate Event, Interrupting

Timer Intermediate Event, Non-Interrupting

All 4 of these Timer Events are “Catching Events” meaning they are waiting for a trigger before emitting a token and allowing flow to proceed down a particular path.

So how does each Timer Event work and under what conditions are they used?

A Timer event can represent a specific point in time (Each morning at 9am, the 30th of the month at 10pm, etc.) or it can be used to represent a time span or passage of time (5 mins, 3 hours, 15 days, etc.)

Timer Start Events (represented as a timer inside a single thin circle) kick off a process once the Timer Event is triggered while Timer Intermediate Events are used within a process that is already underway to redirect the flow of a process.

An interrupting event is more common that a non-interrupting event. As the name implies, once triggered, the current task stops and the flow of control continues along the new path. This is shown in the diagram below.

In the case of an Non-Interrupting Timer Event (represented as a timer inside one or two dashed circles) subprocess A will not be interrupted. Subprocess A will complete and the flow will continue on to subprocess B while a parallel flow will continue from the Non-Interrupting Timer Event onto the Handle Timeout subprocess.

For more details on the Business Process Modeling Notation (BPMN) including BPMN Timer Events, please reference our Cheat Sheet for BPMN (Business Process Modeling Notation).

What are multi-instance activities in BPMN and how are they used?

Chris Adams — Fri, 05 Apr 2013 22:11:00 GMT

Multi-Instance Activities or Tasks in BPMN are one of several forms of looping activities. While looping can always be modeled in BPMN using a decision gateway and sending the flow backwards to an earlier point in the process, a multi-instance activity is a nice shorthand notation that can be used in certain situations.

Multi-instance activities are denoted by three parallel lines at the bottom-center of the activity/task symbol. It’s purpose is to show that the activity occurs for a collection of objects or items. The number of times that the activity completes is defined by the number of items that exist in the collection. Since the collection is delivered to the activity at the start, the number of repetitions is immediately know. This is different than other looping mechanisms that must check a condition every time the loop completes in order to determine if it should continue looping.

Three vertical lines indicate that the multi-instance activity is non-sequential. This means that the activity can be completed for each item in the collection in no particular order.

Three horizontal lines indicate that the multi-instance activity is sequential. This means that the activity must complete for each item in the order that they are received within the collection.

What is a BPMN Collaboration Diagram?

Chris Adams — Mon, 11 Mar 2013 02:30:00 GMT

BPMN Collaboration Diagrams show how two or more processes without a central control interact with each other in synchronized way (how they collaborate).

The term “central control” is key to understanding how to represent processes in a collaboration diagram. In BMPN, a Pool is used to contain and represent a process that is centrally controlled. However sometimes one process interacts with another process while each process is controlled by a different governing body. This commonly happens when one organization’s process is dependent upon information or a delivery of goods from an external organization’s process. When this happens, each process is shown in its own pool, and the two processes can communicate with one another via Message Flows between pools. It should be noted that message flows are only used to communicate between pools and not within a pool.

Message Flows are shown as dashed lines with an empty circle showing where the message originates and and empty arrowhead where the message terminates. In contrast, Sequence flows show how the control is directed from one activity to another and are shown as solid lines with filled in arrowheads. Sequence Flows between activities never cross between pools, since the control of one process does not govern the control of the other process . You can think of the boundary between pools as an interface in which the messages and their sequence are known, but the internal workings of the other process does not need to be known.

When only a single pool exists, it is often labeled to reflect the name of the process. However, within a collaboration diagram where multiple pools exist, the label is used to show the participant which manages the control of the process represented within the pool.

When multiple pools are shown, it is often the case that the organization that owns one process doesn’t know the inner workings of the other organization’s process. So, a pool will be shown but the process within it will not. Instead, only message flows will be shown between the pools.

What do the terms private process and public process mean?

Chris Adams — Mon, 11 Mar 2013 02:30:00 GMT

Processes can be modeled at different levels of abstraction. The decision of what amount of detail to show depends on the audience. A private process typically contains a much greater level of detail and reflects all of the information needed for someone internal to the organization to understand and follow the process. In contrast, a public process will usually contain much less detail than a private process and focuses on showing only the activities and portions of the process that are relevant to other participants in the process.

A great way to visualize this is to consider a process representing a buyer/seller relationship. The buyer and seller are participants in the process, each contained within their own pool since they control their own process. However, they interact with each other via message flows.

If the buyer sends an order to the seller, the seller must fulfill the order. There may be an elaborate quality control check as part of the order fulfillment process. But this detail is irrelevant to the buyer. So in a public process it wouldn’t be shown. In fact, by using public processes (higher levels of abstraction showing less detail) the model allows for the flexibility of each of the participants to change the internal structure of their process without impacting each other. The agreed upon handshake or interface, which is documented by a sequence of message flows, remains unchanged.

What are Message Flows in BPMN?

Chris Adams — Mon, 11 Mar 2013 02:24:00 GMT

What are Message Flows in BPMN?Message flows are used in BPMN collaboration diagrams. Collaboration diagrams show how two or more processes without a central control interact with each other in synchronized way (how they collaborate). Each process is contained within its own pool and the pool is typically labeled with the participants name.

Message flows are how the two separately controlled processes communicate and collaborate with one another. An activity or event in one pool can initiate a message to another pool. Message Flows are depicted as dashed lines with an empty circle showing where the message originates and and empty arrowhead where the message terminates.

When using message flows, the message itself can be depicted as an envelope on the flow line or attached to it by a dotted line. The message represents an item which is transferred, manipulated, or stored during the process. These can be physical items or informational. The message icon (the envelope) can be show unfilled, or with a gray fill. The unfilled icon represents an initiating message while the filled icon represents a non-initiating message (or a message which a participant has sent in response to a previous message it received)

Message Flows can cross the boundaries of a pool, and represent the exchange of information between participants, whereas sequence flows will never cross over the boundaries of a pool. Message flows can be attached to a pool, activity or event.

What is a BPMN Event?

Chris Adams — Mon, 10 Oct 2011 01:26:00 GMT

An Event is a BPMN process modeling element representing something that “happens” during the course of a process. Events affect the flow of the process. Some events such as most start events and some intermediate events have a “trigger” which defines the cause for the event, while end events often define the “result” or consequence of the ending of a particular sequenced flow.

Start Events can only “catch” a trigger. Similarly, End Events can only create “throw” a result. Intermediate events are less constrained and may catch triggers or throw results.

The symbol used for an event is always an unfilled circle. However, there are several other standard attributes of the event symbols that are of importance.

Internal Markers - many times the circle will have an icon or “marker” within it providing more information about the type of event.

Filled versus Unfilled Markers - the internal marker may be filled or unfilled. Unfilled internal markers are used to identify that the event is a “catching” event. This means that the event is waiting to receive notification that something has happened. Filled internal markers are used to identify that the event is a “throwing” event. This means that the event is being generated at that point in the process.

Single, Double, or Filled Outer Circle – the outer border of the event may be a single circle, a double circle, or a filled ring. A single thick circle means that the event is a starting event, used to signify a start of a process. A double circle means that the event is an intermediate event, used for any event which occurs after the start of a process but before its completion. A filled ring means that the event is an end event, used to signify the end of a process.

Dashed versus Solid Circle – both start and intermediate events can be show with the outer circles being solid lines or dashed. Solid lines are used to show interrupting events. These are events which are triggered the moment the event occurs regardless of whether the activity has completed. Dashed lines are used to show non-interrupting events. These are events which will wait for the completion of the activity before flow of the process continues.

For more details on the Business Process Modeling Notation (BPMN) including BPMN Events, please reference our Cheat Sheet for BPMN (Business Process Modeling Notation).

What is a BPMN Gateway?

Chris Adams — Mon, 10 Oct 2011 01:16:00 GMT

A Gateway is a BPMN process modeling element used to control how sequence flows interact as they converge and diverge within a process. The term gateway is used to signify a mechanism by which something is either allowed or disallowed through. A gateway is represented as a diamond. Unlike activities, gateways do not represent work being done within a process flow. They have no effect on operational measures such as cost, time, etc.

When discussing flow control, tokens that traverse the various flow paths are often discussed. Gateways have the ability to consume or generate additional tokens. How these tokens are generated or consumed depends upon the type of gateway used, as identified by its internal marker. The types of internal markers that impact flow control are shown below.

Can you depict data items in a business process diagram?

Chris Adams — Sat, 10 Sep 2011 23:11:00 GMT

Data objects and associations can be depicted in business process flow diagrams. BPMN for example defines a specific notation for dealing with data inputs and outputs from activities. The diagram below shows the standard notation used to show a data object which is generated as output from one activity and used as input by another activity.

The BMPN Assocation connector (dotted line with an arrowhead showing the direction of flow), is used to show the data association between the activity which produces the output and the activity which receives and uses the input. If the data within the data object is defined as required, then the activity that relies on the data object as an input cannot proceed until the data is available.

The following diagram shows another allowed “shortcut” notation that can be used when the data object is an output of one activity and an input into the next activity in sequence. However, data objects do not always have to be used as input to the next activity in sequence. They can be used by activity anywhere later in the process.

What types of business modeling are supported by BPMN 2.0?

Chris Adams — Sat, 10 Sep 2011 23:09:00 GMT

BPMN is intended to support only business process modeling. This means that other important types of business modeling are not within the scope of the BPMN standard.

Types of business modeling not supported by BPMN include:

Definition of organizational models and resources

Modeling of functional breakdowns

Data and information models

Modeling of strategy

Business rules models

What is the difference between a Pool and a Swimlane?

Chris Adams — Sun, 08 May 2011 02:06:00 GMT

A “swimlane” is a generic industry term used to describe the grouping of common activities in a process diagram into labeled rectangular area.

Two of the more common standards for process diagrams are the UML Activity Diagram and the BPMN Business Process Diagram (aka Swimlane Diagram).

In UML, the generic “swimlane” is represented by a concept that is called an Activity Partition. The UML 2.0 specification describes an activity partition as a “kind of activity group for identifying actions that have some characteristic in common.” Activity Partitions can be further divided into Activity Sub-partitions. There is really no difference between an Activity Partition and an Activity Sub-Partition other than the level at which the group activities based on a common set of characteristics.

In BPMN, the generic term “swimlane” is represented by either a Pool or a Lane in the swimlane diagram. A Pool is a graphical representation of a participant within the business process diagram. Pools can represent different businesses, a generic business partner (like a supplier, manufacturer, or buyer), or be more granular if modeling a single business without external dependencies. Additionally, a Pool may reference a process and show the internal details (white box), or it may have no internal details (black box). A Pool acts as a container for the activities and sequence flows between activities.

Lanes are mentioned in the BPMN specification as being used to organize and categorize activities within a pool. However, the precise meaning of a lane or its appropriate scope is left up to the modeler, as the specification is silent in this area.

So, besides the fact that a Lane is used to subdivide a Pool, what is ultimately the difference between the two? Sequence flows can cross Lanes between activities but they cannot cross Pools. Only messages can pass between Pools. This is by design and is a major benefit of the BPMN notation. Messages indicate transfer of information within a conversation between two parties. These are points of risk within the business process and deserve extra attention. Both minimizing points of communication as well as ensuring clear communications take place help reduce process risk.

Additionally, participants of one pull have no control over the completion of activities that may occur in another pool. Whereas, activities divided between lanes can sometimes be performed by the same person, even if they that person is performing a different role (wearing multiple hats).

What is a Swimlane Diagram?

Chris Adams — Sat, 09 Apr 2011 22:42:00 GMT

A swimlane diagram (also sometime called a cross-functional diagram) documents the steps or activities of a process flow or workflow. More specifically, a swimlane diagram groups these activities into swimlanes which are horizontal or vertical columns that contain all of the activities which fit into the category represented by that swimlane. Swimlanes can represent many categories of information such as actors which perform the activities (i.e., role or department), the stage of the process in which the activity takes place, or whatever else the creator of the document feels should be emphasized and communicated by the swimlane diagram. The term swimlane was adopted due to the visual similarity between the horizontal rows of the diagram to that of the swimlanes found within a swiming pool.

A process flow or workflow diagram does not have to use swimlanes. However, since swimlanes communicate additional information about who performs the activity or when it takes place, it’s typically a preferred best practice to include them. Similarly, a swimlane diagram can use only one set of swimlanes (either vertical swimlanes or horizontal swimlanes). In BPMN standards, a process flow is typically shown flowing from left to right. If only one set of swimlanes is used, horizontal swimlanes are most commonly used to group activities. In the UML standard, the activity diagram flows from top to bottom and veritcal swimlanes are most commonly used.

One of the more common choices used by creators of swimlane diagrams is to define the roles which perform each activity within horizontal swimlanes and define the process stages in which the activity occurs within vertical swimlanes.

Name a few of the industry standards, methodologies, or best practices used by business analysts and systems analysts?

Chris Adams — Thu, 20 Sep 2007 00:57:00 GMT

-UML

-BPMN

-RUP