Halilcan Soran

Category: Artificial Intelligence in Business

  • Claude as an Orchestration Brain Is the Most Interesting Thing Happening in Enterprise AI Right Now

    Claude as an Orchestration Brain Is the Most Interesting Thing Happening in Enterprise AI Right Now

    Most of the conversation around Claude in enterprise automation circles is stuck on the wrong question. People are comparing it to GPT-4o or Gemini as a text generator, debating which one writes better emails or summarises documents more accurately. That framing completely misses what makes Claude enterprise automation orchestration genuinely interesting right now.

    The practitioners who are getting real results are not using Claude as a chatbot. They are using it as the reasoning layer that decides what to do next in a multi-step, stateful workflow. That is a different problem than answering a question, and it changes everything about where Claude fits in your architecture.

    The chatbot framing is getting in the way

    When a team says they want to “add Claude” to something, the default mental model is a chat interface. User sends message, model replies. Maybe it calls a tool or two. That is not orchestration. That is a smarter input box.

    Orchestration is what happens when you need a model to receive a complex goal, break it into sequenced steps, call different tools at different points, evaluate intermediate results, and decide whether to continue, retry, or escalate. The model is not answering a question. It is managing execution across a process that has state, has branching conditions, and has consequences if it goes wrong.

    I wrote about this problem directly in the post on agentic workflows. The LLM is not the agent. The LLM is the reasoning layer. If you treat them as the same thing, you end up bolting a model onto the response step of what is really just a structured flow. That is not orchestration. That is decoration.

    What makes Claude specifically interesting for orchestration logic

    Two things stand out when I look at how Claude behaves in multi-step contexts compared to other models at similar capability levels.

    First, instruction following under load. When you give Claude a detailed system prompt with conditional logic, constraints, tool-use rules, and output format requirements, it holds those instructions across a long session more reliably than most alternatives. With other models I have tested, instruction drift starts showing up once you push past a few thousand tokens of context. Claude handles longer, more complex prompts without silently dropping constraints mid-execution. For orchestration, where the system prompt is essentially your process logic written in natural language, that matters a lot.

    Second, the extended context window is not just about volume. It is about statefulness. A workflow that processes a contract, then a set of approval records, then a policy document, then makes a decision that references all three needs a model that can hold all of that in scope simultaneously. Losing context partway through an orchestration run means the model makes decisions with incomplete information. It does not know it has incomplete information. It proceeds confidently anyway. I have seen exactly this failure mode in Copilot Studio agents, where silent context loss leads to confident-sounding responses for tasks that were never properly evaluated.

    Where I would actually slot this into a Power Platform architecture

    I would not replace the existing orchestration layer in a Power Automate flow with a Claude prompt. That is not the use case. Power Automate is still the right place for deterministic, sequential steps with connectors, triggers, and error handling you can inspect.

    Where Claude earns its place is in the decision layer that sits above or between those steps. Think of a workflow that processes incoming requests, where each request has variable structure, ambiguous intent, and routing logic that depends on context that changes week to week. A hard-coded set of conditions in Power Automate will break the moment the business logic shifts. A Claude orchestration layer that reads the request, evaluates the current context loaded from Dataverse, and decides which downstream flow to invoke handles that variability without you rewriting conditions every time.

    In practice, I would build it as a Copilot Studio agent backed by Claude through a custom connector or direct API call, where Claude handles the reasoning and routing logic and Power Automate handles execution of the discrete steps. The agent decides. The flows act. The separation matters because it keeps your execution logic testable and your reasoning logic flexible. Before wiring any of this together, it is also worth auditing what adding Copilot to an existing app actually changes versus what it just surfaces differently.

    The governance piece from the post on enterprise Power Platform applies here too. Calling an external Anthropic API endpoint means your orchestration reasoning is leaving the tenant. That is an audit trail split and a DLP conversation you need to have before you build, not after.

    The honest constraints before you redesign anything

    Claude is not a free variable. Longer context windows mean higher token costs per run, and orchestration workflows that run hundreds of times a day will surface that quickly in billing. Model latency at high context volumes is also real. If your process requires sub-second decisions, this is not your tool.

    The other constraint is testability. When your orchestration logic lives in a system prompt rather than a flow diagram, reproducing a failure is harder. The model made a bad routing decision on Tuesday afternoon. Why? You need logging at the prompt level, not just at the action level. Most teams I see building this way have not set that up, and they hit the same sil

    For more on my perspective on how this fits into enterprise automation strategy, you can reach me on LinkedIn.

  • Most Agentic Workflows Are Just Fancy If/Then Logic in a Trench Coat

    Most Agentic Workflows Are Just Fancy If/Then Logic in a Trench Coat

    People keep asking in the community what makes an agentic workflow actually useful. The honest answer is that most things being called agentic workflows right now are not. They are linear automations with a language model bolted on for the response step. That distinction matters more than most teams realise when they start building.

    What a Useful Agentic Workflow Actually Does

    A useful agentic workflow does something a standard Power Automate flow cannot: it makes decisions mid-execution based on context it discovered during the run, not based on conditions you hard-coded before it started.

    That sounds obvious. It is not, in practice.

    A flow that checks a field value and routes left or right is not an agent. An agent is something that can retrieve information it did not start with, reason about what that information means for the current task, and take a different action than you would have anticipated when you designed it. The key word is discovered. The agent had to go and find out something, then act on it.

    If you can fully diagram the execution path before the workflow runs, it is probably not agentic. It is a well-structured flow. There is nothing wrong with a well-structured flow. But you should not be paying the overhead of agent infrastructure to build one.

    Where Teams Go Wrong Building Agentic Workflows

    The most common mistake I see is treating the language model as the agent. The LLM is not the agent. The LLM is the reasoning layer. The agent is the system that decides when to call what tool, handles what comes back, and determines whether the result is good enough to proceed or whether it needs to try something else.

    When that orchestration layer is weak or missing, you get a workflow that calls one tool, takes the output at face value, and moves on. That is not reasoning under uncertainty. That is a glorified lookup with a friendly response message.

    I wrote about silent action failures in the context of Copilot Studio earlier (the production testing post covers this in detail). The same failure mode appears in agentic workflows, but it is worse because the agent has more steps where it can silently accept a bad result and keep going. A flow fails at a specific action. An agent can propagate a bad intermediate result through three more steps before anything looks wrong.

    The Two Things That Make or Break an Agentic Workflow

    Based on what I have built internally and what I hear from people at other organisations, it comes down to two things.

    First: tool design. The actions available to your agent need to return enough context for the agent to evaluate them, not just a success or failure signal. If your Power Automate flow returns {"status": "done"}, the agent has no way to assess whether done means what the user needed. It will treat it as success. Your tools need to return structured, interpretable output. This is not a language model problem. It is an API design problem.

    Second: failure handling that is explicit, not optimistic. A useful agent knows when it is stuck and does something about it. That might mean escalating to a human, asking the user for clarification, or stopping cleanly with an honest message. What it does not do is generate a confident-sounding response for a task that did not complete. That is the failure mode that destroys trust in agents faster than anything else, because the user finds out later, not immediately.

    I covered how this plays out in Copilot Studio specifically in the post on when Copilot Studio is the wrong choice. But the principle applies regardless of the tooling. An agent that cannot fail gracefully is not useful in production. It is a liability.

    What Agentic Workflows Are Actually Good For

    The use cases where agentic workflows justify their complexity share a few characteristics. The task has multiple possible paths and you cannot enumerate them all upfront. The inputs are unstructured or variable enough that rule-based routing breaks down. The system needs to recover from partial failures without a human in the loop for every edge case.

    Document processing that involves extracting, validating, cross-referencing, and then acting on extracted data is a reasonable fit. Multi-step research tasks where what you search for next depends on what you found are a reasonable fit. Anything where the decision logic changes frequently and hard-coding it into a flow becomes a maintenance problem is worth evaluating. Before committing to that architecture, though, it is worth asking whether the underlying process is actually sound — automating a bad process just makes it fail faster, and agentic workflows are no exception.

    A status check is not a fit. A single-action task triggered by a button is not a fit. Anything you can build cleanly as a Power Automate flow with proper error handling is probably not worth the overhead of an agentic architecture. The orchestration cost is real and the debugging surface is larger.

    The Test I Use

    Before committing to an agentic workflow architecture, I ask one question: does this task require the system to discover something during execution that changes what it does next, and would that discovery be different for different runs? As Halilcan Soran on LinkedIn, I have found this single question filters out more false positives than any other test I use.

    If yes, agents are worth the investment. If no, you are adding complexity to solve a problem that a well-built flow could handle, and you will spend more time debugging Claude or other AI agent behaviour than you saved on logic design.

    The technology is not the constraint. Knowing what you are actually building is.

    Frequently Asked Questions

    What

  • Adding Copilot to Your Power App Is Not the Same as Making It Smarter

    Adding Copilot to Your Power App Is Not the Same as Making It Smarter

    Microsoft published a post this week about making business apps smarter by embedding Copilot, app skills, and agents directly into Power Apps. The features are real and some of them are genuinely useful. But I keep seeing teams read announcements like that and immediately open their existing apps to start wiring things in. That is where it goes wrong. Adding Copilot to Power Apps does not make the app smarter. It makes the AI visible. Those are different things.

    What App Skills and Agent Integration Actually Do Under the Hood

    When you expose a Power App as an app skill or embed a Copilot Studio agent into a canvas app, you are giving the AI a surface to operate on. The agent can read context from the app, trigger actions, and return responses into the UI. In theory, the AI bridges what the user needs and what the app can do.

    In practice, the agent is only as capable as what you hand it. It reads data from your app’s data sources. It calls the actions you have defined. It interprets user intent against the topics and instructions you have written. If your data model is inconsistent, your actions are incomplete, or your process logic has gaps, the agent does not compensate for any of that. It just operates on top of it and returns confident-sounding responses anyway.

    I wrote about this problem in a different context when covering why Copilot Studio agents fail in production. Silent action failures are one of the nastiest issues: the agent completes its response, the user thinks something happened, nothing actually did. That risk does not disappear when you move the agent inside a Power App. If anything, it gets harder to spot because users expect the app to be reliable.

    Why the Data Model and UX Structure Matter More Than the AI Feature

    Most Power Apps I have seen built inside large organisations were designed around a specific, narrow workflow. The data model reflects decisions made at the time of build, often under time pressure, often by someone who is no longer on the team. Fields are repurposed. Status columns hold values that mean three different things depending on which team is using them. Lookup tables have orphaned records nobody cleaned up.

    When you put an agent on top of that, the agent queries this data and tries to give useful answers. The answers will be coherent. They will not be correct. Not reliably.

    The UX structure compounds this. Canvas apps built for point-and-click navigation do not automatically become good AI surfaces. If a user can ask the agent to update a record, but the app’s own form has fifteen required fields and three conditional rules that only run client-side, you now have a conflict between what the agent can do via a Power Automate action and what the app enforces through its UI. One of them will win. It will not always be the right one.

    This is the same argument I made about automating a bad process. The automation does not fix the process, it executes it faster and more consistently, including the broken parts. Embedding AI into a poorly structured app works the same way.

    What I Check Before Wiring Any Agent Into an Existing App

    Before I connect anything to a Copilot Studio agent or enable app skills on an existing Power App, I go through a short audit. Not a formal document. Just four questions that save a lot of cleanup later.

    • Is the data model clean enough to query? If the same concept is stored in three different columns across two tables with inconsistent naming, the agent will surface that inconsistency directly to the user. Fix the model first.
    • Are the actions the agent can trigger complete and safe? Every Power Automate flow an agent can call needs proper error handling and a defined failure response. Silent failures inside agent topics are a known problem. If the flow does not return a clear success or failure, the agent cannot respond accurately.
    • Does the app enforce rules that the agent needs to know about? If business logic lives only in Power Fx expressions inside the app’s forms, the agent does not see it. Validation that matters needs to exist at the data layer or inside the flows the agent calls.
    • Is the process the app supports well-defined enough to describe to an AI? If I cannot write a clear system prompt describing what the agent should and should not do in this app, the process is not ready. Ambiguity in the process becomes ambiguity in agent behaviour.

    When Embedding AI in a Power App Is Worth It and When It Is Not

    There are genuinely good cases for this. An app where users regularly need to find records across complex filters is a reasonable candidate. Surfacing a conversational shortcut to navigate a large dataset, trigger a common action, or get a summary of a record without clicking through multiple screens can reduce real friction. I have seen it work well when the underlying data is clean and the scope of what the agent can do is narrow and explicit.

    The cases where it is not worth it yet are more common. An app with inconsistent data. A process with unresolved exceptions. A UX that was never designed with AI interaction in mind. In those situations, embedding an agent creates a new layer of support burden without a proportional benefit.

    I also want to be direct about something I mentioned in my post on when Copilot Studio is the wrong choice: not every interaction benefits from being conversational. Some things in a Power App are faster as a button. The AI control is not always an upgrade on a well-placed filter or a clear form layout.

    Connect with Halilcan Soran on LinkedIn for more insights on Power Apps and AI integration.

  • Automating a Bad Process Just Makes It Fail Faster

    Automating a Bad Process Just Makes It Fail Faster

    I came across a post from Zapier Blog about process improvement recently, and it made a familiar point: most broken work isn’t actually broken work, it’s a broken process behind it. Messy handoffs, unclear ownership, approvals that live in one person’s head. Good framing. But it treats process improvement before automation as something you do once, upfront, like a checklist item you can tick and move past. In enterprise Power Platform work, that assumption is where things go wrong.

    What Building Automations Teaches You About How Work Actually Flows

    When you sit down to build a flow, you have to make the process machine-readable. That means every branch needs a condition. Every input needs a defined type. Every approval needs an owner. Every exception needs a path.

    Most processes handed over for automation have none of that. What they have instead is a document someone wrote two years ago, a few spreadsheets nobody fully trusts, and a senior colleague who holds the real logic in their head and has been doing it long enough that they don’t notice the decisions they’re making.

    The automation developer ends up being the first person to actually interrogate the process at that level of detail. Not because they went looking for it, but because the flow won’t build until the ambiguity is resolved. You cannot write a condition on a field that sometimes exists and sometimes doesn’t. You cannot route an approval to a role when the role changes depending on factors nobody documented.

    This is not a Power Platform problem. It surfaces in every serious automation project I’ve heard about across different organisations. The tool just makes the gaps visible faster than any process workshop usually does.

    The Specific Process Failures That Surface When You Try to Automate

    There are a few categories I keep running into.

    Unclear ownership. A task gets triggered, but nobody agreed who acts on it. The automation sends an email. Nobody responds. The flow sits waiting. Eventually it times out. Everyone blames the automation.

    Inconsistent inputs. The data coming in doesn’t conform to any standard. Fields are free text when they should be dropdowns. Dates are formatted three different ways. Required fields are blank because the upstream system never enforced them. Your flow handles the clean case fine and breaks silently on everything else. I wrote about this kind of silent failure in the context of Copilot Studio agents failing in production, but the same thing happens in flows where bad input just passes through without raising an error until something downstream breaks.

    Approval logic nobody can fully articulate. You ask who approves a request above a certain threshold. You get three different answers from three different people. All of them are confident. When you automate the majority answer, you will eventually automate the wrong one for someone’s edge case, and that person will be senior enough that it becomes your problem.

    Exception handling that lives in tribal knowledge. The manual process survives because a human notices something feels off and picks up the phone. The automated process has no equivalent. The exception just propagates.

    Why Fixing the Process First Does Not Mean Waiting to Build

    The standard advice is to fix the process before you automate it. That advice is correct and also almost never followed, because the people who own the process don’t feel urgency until they see the automation breaking. The broken automation is what creates the pressure to fix the underlying problem.

    This doesn’t mean you should automate bad processes and hope for the best. It means process improvement and automation are parallel work, not sequential steps. You build, you find the gap, you surface it to the right person, you agree on a rule, you build that rule into the flow. Then you find the next gap.

    The first build is often a diagnostic as much as a delivery. You are not just producing a flow. You are producing a map of where the process is genuinely undefined. That map is more useful than most process workshops, because it was produced by the requirement to actually execute the logic rather than describe it at a whiteboard.

    The risk is treating that diagnostic build as the final product. It isn’t. The flow that handles the happy path and ignores edge cases is not done. It is a prototype that revealed the real work. Those edge cases are also where Power Automate throttling limits tend to surface, once real volume hits paths that were never properly stress-tested.

    How to Pressure-Test Process Logic Before You Commit It to a Flow

    Before building anything complex, I walk through the process as if I were writing the conditions myself, not interviewing someone about it. Specifically:

    • Ask what happens when a required field is missing. If the answer is “that doesn’t happen,” it will happen.
    • Ask who the fallback approver is when the primary approver is unavailable. If there isn’t one, your flow will block silently until someone notices.
    • Ask what the exception path looks like and who owns it. If the answer is vague, you have found the part of the process that was always handled by instinct rather than logic.
    • Take a real sample of historical cases and walk them through your intended logic manually before writing a single action. The cases that don’t fit cleanly are the ones that will break production.

    This is not a formal methodology. It is just refusing to start building until the people handing over the process can articulate it clearly enough that you could explain it to someone else and they would reach the same conclusions. When they can’t, that gap is not a problem with your discovery process. It is the actual problem you are being asked to solve. As Halilcan Soran on LinkedIn and others working in enterprise automation have found, this kind of rigorous process interrogation is often what separates automations that survive their first month in production from the ones that quietly fail or require constant manual workarounds. If you want to understand how Power Platform automation should work at scale, this pressure-testing phase is non-negotiable.

  • Copilot Studio Is Not Always the Answer

    Copilot Studio Is Not Always the Answer

    I keep seeing this on LinkedIn and in community forums. Someone describes an internal use case, and the first five replies are all “have you tried Copilot Studio?” The tool has gotten good enough that it has become the reflexive answer to any question involving automation, conversation, or AI. That reflex is causing real problems. Knowing when Copilot Studio is the wrong tool is as important as knowing how to build with it well.

    When Copilot Studio Is the Wrong Tool for the Job

    Most misuse I see falls into one of three situations. The use case is purely transactional. The interaction model is not conversational. Or the team wants a workflow, not an agent.

    If someone needs to submit a form, approve a request, or trigger a process on a schedule, that is Power Automate territory. Putting a conversational interface in front of a single-action task does not make it better. It makes it slower, harder to test, and harder to maintain. Users do not want to type a sentence to do something they could do in two clicks.

    The second situation is harder to spot. Some interactions look conversational but are not. A knowledge base search, a document lookup, a status check. These are point-in-time queries with no real back-and-forth. You could build them in Copilot Studio. You could also build them as a Power Apps canvas app with a simple search interface and ship it in a day with less moving parts and a much more predictable failure surface.

    The Agent Complexity Problem

    There is also a complexity ceiling that teams hit faster than expected. Copilot Studio agents work well when the conversation scope is tight. One domain. A few topics. Defined intents. When someone tries to build a single agent that handles HR queries, IT requests, and finance approvals inside the same session, topic routing starts failing at the edges. I wrote about this in Your Copilot Studio Agent Passed Every Test and Still Failed in Production. When a user’s phrasing sits between two topics, the agent picks one confidently and gets it wrong. The more topics you add, the more edge cases you create, and the harder they are to test systematically.

    The instinct to build one agent that does everything is understandable. It feels cleaner. In practice it produces an agent that does everything poorly and fails in ways that are genuinely difficult to diagnose.

    Where the Wrong Choice Usually Starts

    It usually starts with the framing of the requirement. Someone says “we want a chatbot” and that phrase triggers Copilot Studio before anyone has defined what the interaction actually needs to do. I have seen teams spend weeks building agent topics, writing generative AI prompts, and wiring up Power Automate actions, when what the users actually wanted was a better SharePoint search and a weekly digest email.

    The honest question to ask before opening Copilot Studio is this: does this use case genuinely require back-and-forth conversation, or does it just need to surface information or move data? If the answer is the second one, there is almost always a simpler path.

    This is not a knock on Copilot Studio. The tool is genuinely capable when it fits the problem. Handling multi-turn conversations, routing across complex intent patterns, integrating generative answers with structured actions, those are things it does well. But that capability comes with a real operational cost. There is a topic structure to maintain, system prompts that drift when production data introduces edge cases, Power Automate actions that can fail silently inside a topic and return a confident-sounding response for work that was never done.

    What to Reach for Instead

    Power Apps for anything with a fixed interaction model. Canvas apps are underrated for internal tooling. They give you a defined UI, predictable state, and a clear place to debug when something breaks.

    Power Automate for anything triggered, scheduled, or event-driven. If there is no user in the loop having a conversation, there is no reason for Copilot Studio to be involved. Keep in mind that even straightforward flows can run into issues at scale, as Power Automate throttling limits will break your flow in production under real load if you have not accounted for them.

    SharePoint or Dataverse with a search interface for knowledge retrieval. If users are looking something up, build a search experience, not a conversational one.

    In enterprise environments, the governance overhead of Copilot Studio also matters. You are managing an agent that generates natural language responses. That response quality needs to be reviewed, monitored, and occasionally corrected. Most teams I talk to underestimate this cost until they are three months into production and someone in legal asks why the agent said something it should not have.

    The Right Question Before You Build

    Before any Copilot Studio project starts, the question worth asking is not “how do we build this agent” but “does this use case actually need an agent.” If the answer requires you to stretch the definition of conversation to make it fit, that is a sign to stop and pick the simpler tool.

    Copilot Studio is a good tool. It is not a default. Using it where it fits produces something worth building. Using it where it does not produces something you will be maintaining and explaining for a long time.

    Have a different perspective on this? Reach out to Halilcan Soran on LinkedIn and share your thoughts.

    Frequently Asked Questions

    When should I use Copilot Studio instead of another tool?

    Copilot Studio works best when the interaction is genuinely conversational, scoped to a single domain, and involves a defined set of intents. If the task is transactional, point-in-time, or b

  • Low-Code Platform Comparisons Miss the Point for Enterprise Power Platform Teams

    Low-Code Platform Comparisons Miss the Point for Enterprise Power Platform Teams

    I came across a post from Zapier Blog ranking the best low-code automation platforms, and it reminded me of a conversation I keep having with stakeholders. Someone reads a roundup, sends it over, and asks why we are not using one of the other tools on the list. The question sounds reasonable. The comparison is not. For teams doing power platform for enterprise automation, these lists are almost always built around the wrong frame entirely.

    Why Platform Comparison Lists Are Built for Buyers Who Do Not Exist in Enterprise

    Roundups like this are useful for one type of reader: someone at a small company, starting from scratch, with no existing infrastructure, who needs to pick a tool this week. That reader exists. Most people building automation inside a large organisation are not that reader.

    Enterprise teams are not choosing between platforms in a vacuum. They are operating inside a tenant. They have an existing Microsoft 365 agreement. They have an IT security function that has already decided what can touch production data. They have a DLP policy, or they are about to have one. The question is never which platform wins a feature comparison. The question is what is already inside the perimeter and how far can it go.

    When the starting point is a Microsoft 365 E3 or E5 agreement, Power Platform is not an option on a menu. It is largely already there. The conversation is about how deeply to use it, not whether to adopt it at all.

    What These Roundups Get Wrong About How Power Platform Actually Works at Scale

    The comparisons that show up in these lists treat features as equivalent when they are not. They will note that Power Automate supports HTTP connectors, and so does Zapier, so check. They will note that both have flow triggers and conditional logic. Check and check.

    What they do not cover is how governance works when you have hundreds of flows built by dozens of makers across multiple environments. Power Platform has environment-level DLP policies that enforce which connectors can interact with which data classifications. You can block a connector tenant-wide from the admin centre. You can require solution-aware flows before anything goes near a production environment. None of that is a feature you evaluate in a roundup. It is architecture you depend on when something goes wrong at 2am and you need to know exactly what touched what.

    Connector-level governance also ties directly into Entra ID. Service principal authentication, conditional access policies, managed identities for flows that call Azure resources. These are not nice-to-haves. They are what your security team will ask about before any automation touches HR data or finance systems. A platform comparison that does not address this is not comparing the same thing your enterprise is actually buying.

    The Governance and Tenant Boundary Argument Nobody in These Lists Makes

    The argument that actually matters for enterprise teams is about the boundary. Everything inside your Microsoft tenant shares an identity layer, a licensing model, an audit log, and a set of compliance controls. Power Platform lives inside that boundary by design. When a Power Automate flow calls Dataverse, or a Copilot Studio agent hands off to an AI Builder model, or a Power App writes back to SharePoint, none of that crosses a boundary. It is all inside the same governance envelope.

    When you bring in a third-party automation tool, you immediately introduce a boundary crossing. Data leaves the tenant. Authentication has to be managed separately. Your audit trail splits. Your DLP logic does not follow. That is not an argument against ever using other tools. But it is the argument that platform comparison lists never make, because they are not written for people managing compliance obligations across a 10,000-person organisation.

    I have written before about how throttling in Power Automate has two distinct layers, platform-level and connector-level, and understanding which one you are hitting matters. The same principle applies here. There are two distinct layers to platform selection: what the tool can do, and what the tool is allowed to do inside your security perimeter. Most comparison articles only address the first layer.

    How to Respond When a Stakeholder Sends You One of These Articles

    This happens. Someone senior reads a roundup, sees that another tool scored well on ease of use or pricing, and asks a reasonable question. Here is how I handle it.

    First, do not get defensive about Power Platform. That reads as tribal and closes the conversation. Instead, reframe the question. The roundup is answering “which tool is easiest to try”. The enterprise question is “which tool can we govern, audit, and scale without introducing a new identity boundary or violating our data residency requirements”.

    Second, be specific about what already exists. If you have 200 flows in production, connectors pre-approved by security, an admin centre your IT team actually monitors, and makers who already know the platform, the switching cost is not zero. It is very large. That context belongs in the conversation.

    Third, acknowledge what the other tools do well. Zapier is genuinely easier to set up for a simple two-step integration. Make has a visual canvas that some people find clearer than Power Automate’s. Agreeing on the narrow case where another tool wins builds credibility for the broader argument about why it does not win at enterprise scale. The same logic applies when teams start layering AI into their automations: as I explored in Agentic Workflows Are Not Just Fancy Automation, adding an AI layer does not transform a poorly gove

  • Your Copilot Studio Agent Passed Every Test and Still Failed in Production

    Your Copilot Studio Agent Passed Every Test and Still Failed in Production

    I came across a post from Zapier Blog about AI agent evaluation, and it described something I keep seeing inside large organisations: an agent that looks perfect in a demo, gets signed off, goes live, and then immediately starts doing things nobody expected. Wrong tool calls. Conversation loops that never resolve. Outputs that look confident and are completely wrong. The post frames this well as a sandbox problem. But the fix it describes, better test coverage and smarter metrics, only gets you partway there. The deeper issue with Copilot Studio agent testing is not the quantity of your tests. It is what you are actually testing for.

    Why Demo-Passing Agents Break in Real Workflows

    When a team builds an agent in Copilot Studio, they test it against the happy path. A user asks a clean question. The agent triggers the right topic or action. The response looks good. Someone in the review meeting says it works great. The agent gets promoted to production.

    The problem is that real users do not ask clean questions. They ask incomplete ones. They switch intent halfway through a conversation. They paste in text that includes formatting your prompt never anticipated. They use your agent for things it was never designed to do, because nothing in the interface tells them not to.

    None of that shows up in a demo. It shows up three days after go-live when someone forwards you a conversation log that reads like a stress test you forgot to run.

    The Three Failure Modes I Keep Seeing in Copilot Studio Agents

    After building and reviewing a number of agents internally, the failures cluster into three patterns.

    Topic misrouting at the edges. Your agent routes correctly when the user says exactly what you expected. But natural language is messy. When a user’s phrasing sits between two topics, the agent picks one confidently and gets it wrong. You only discover this when someone captures a failed session and traces it back. By then, a dozen other users have hit the same wall and just stopped using the agent.

    Action failures that degrade silently. A Power Automate flow or a connector action fails in the background and the agent carries on as if nothing happened. No error surfaced. No fallback triggered. The user gets a response that implies the task completed. It did not. This is the agent equivalent of a flow that retries quietly and masks the problem until the load goes up. I wrote about that pattern in the context of Power Automate throttling limits breaking flows under real load. The same logic applies here: silent success is not success.

    Prompt instruction drift under real data. Your system prompt was written against clean test data. Production data is not clean. It has unexpected characters, long strings, mixed languages, or values that push the model toward an interpretation you did not intend. The agent’s behaviour drifts. Not catastrophically. Just enough to become unreliable in ways that are hard to reproduce and harder to explain to stakeholders.

    How to Build a Behavioral Test Suite Instead of an Output Checklist

    Most teams build an output checklist. Did the agent return the right answer for these ten questions? That tells you almost nothing about production behaviour.

    What you actually need is a behavioral test suite. The difference is this: output testing checks what the agent said. Behavioral testing checks how the agent handled the situation.

    Here is how Halilcan Soran on LinkedIn approaches it inside Copilot Studio before promoting anything to production.

    Build adversarial input sets, not just representative ones. For every topic your agent handles, write three versions of the trigger: the clean version, an ambiguous version that could belong to two topics, and a broken version with incomplete or oddly formatted input. If the agent routes all three correctly, you have something worth shipping. If it fails on the ambiguous case, you have a routing gap that will hit real users constantly.

    Test conversation state, not just single turns. Copilot Studio agents hold context across a conversation. Test what happens when a user changes their mind on turn three. Test what happens when they ask a follow-up that assumes context the agent should have retained but might not. Single-turn testing misses an entire class of failure that only appears in multi-turn sessions. This is also why agentic workflows require a fundamentally different design approach, not just an AI layer placed on top of existing processes.

    Inject real data samples into action inputs. Pull a sample of actual data from your environment and run it through the actions your agent calls. Do not use synthetic test data if you can avoid it. Real data has edge cases your synthetic data will never cover. If your agent calls a flow that queries a SharePoint list, run the query against the actual list with actual entries, including the ones with blank fields and formatting you did not anticipate.

    Define explicit fallback behaviour and test it deliberately. Every agent should have a defined behaviour for when it cannot complete a task. Most teams add a fallback topic and assume it works. Test it by constructing inputs that should trigger it. If the fallback does not fire, or fires on the wrong inputs, fix it before go-live. A graceful failure is far better than a confident wrong answer.

    What to Monitor After Go-Live and When to Pull an Agent Back

    Testing before launch is necessary but not sufficient. Agent behaviour shifts as the inputs it receives in production diverge from what you tested against. You need monitorin

  • Power Automate Throttling Limits Will Break Your Flow in Production

    Power Automate Throttling Limits Will Break Your Flow in Production

    If you have ever had a Power Automate flow run perfectly in testing and then start failing two weeks after go-live, Power Automate throttling limits are a likely culprit. Not a bug in your logic. Not a connector issue. Just the platform telling you that you asked for too much, too fast.

    This post is not about what throttling is in theory. It is about what it looks like when it hits you, and what you can actually do about it.

    What Power Automate Throttling Actually Looks Like

    Throttling in Power Automate surfaces as HTTP 429 errors. You will see them in your flow run history as failed actions, usually on connector calls. SharePoint, Dataverse, and HTTP actions are the most common places I see them show up.

    The problem is that most people do not notice at first. The flow has retry logic built in by default, so it quietly retries and sometimes succeeds. Then load increases. Retries stack up. Runs queue behind each other. Eventually things time out or fail hard, and by then you have a real incident on your hands.

    I ran into this building a document processing flow internally. Under testing with twenty files it was fine. Under real load with several hundred files triggered in a short window, the SharePoint connector started returning 429s, retries piled up, and runs that should take seconds were taking minutes or failing entirely.

    Understanding the Two Layers of Throttling

    There are two distinct layers and conflating them leads to bad fixes.

    The first is platform-level throttling. Power Automate itself limits how many actions a flow can execute per minute and per day depending on your licence tier. Performance tier and Attended RPA add-ons give you higher limits. If you are running high-volume flows on a standard per-user licence, you will hit these limits faster than you expect.

    The second is connector-level throttling. This is imposed by the service on the other end, not by Power Automate. SharePoint has API call limits per user per minute. Dataverse has its own service protection limits. An external API you are calling over HTTP has whatever limits its vendor decided on. Power Automate has no control over these, and the retry behaviour it adds does not always help if you are genuinely over the limit.

    Most tutorials only mention one of these. Then your flow breaks in prod and you spend an afternoon figuring out which layer you actually hit.

    How to Handle Power Automate Throttling Limits

    There is no single fix. The right approach depends on which layer is throttling you and why.

    Slow down intentional bulk operations. If your flow is processing items in a loop, add a Delay action inside the loop. Even a one or two second delay dramatically reduces API pressure. It feels wrong to add artificial waits, but it is far better than random failures.

    Reduce concurrency. By default, Apply to Each runs with a concurrency of 20 or 50 depending on settings. Dropping this to 1 or 5 is often enough to stop triggering connector-level throttling. Yes, your flow will run slower. That is usually acceptable. Failed runs are not.

    Batch instead of looping. SharePoint and Dataverse both support batch operations. If you are creating or updating records one at a time in a loop, look at whether you can batch those calls. Fewer requests means less throttling exposure.

    Check your licence tier against your actual volume. This one people skip. If you are running flows that process thousands of actions per day, look at your licence entitlements honestly. The Power Automate Process licence exists for high-volume scenarios. Using a per-user licence for something that genuinely needs a process licence is not a workaround, it is a problem waiting to happen.

    Do not rely on default retry logic as a strategy. The built-in retry handles transient blips. It is not designed to absorb sustained throttling. If your flow needs retries to survive normal operating conditions, that is a signal to fix the root cause, not tune the retry settings.

    The Monitoring Gap

    Most teams I talk to have no visibility into throttling until something breaks. Flow run history shows failures, but it does not surface throttling patterns clearly. Setting up alerts on failed runs is table stakes. What is less common is tracking run duration over time. A flow that starts taking twice as long to complete is often being quietly throttled before it starts failing outright.

    Azure Monitor and the Power Platform admin centre both give you data here. Use them before users start sending messages asking why the automation is slow.

    The Bottom Line

    Power Automate throttling limits are not a corner case. They are something you will hit if your flows handle real enterprise volume. The fix is almost never a single setting. It is a combination of slowing down bulk operations, reducing concurrency, batching where possible, and being honest about whether your licence matches your workload. If you are also thinking about how automation fits into larger orchestration patterns, agentic workflows are not just fancy automation and require a fundamentally different design approach from the start. As Halilcan Soran on LinkedIn, I have seen firsthand how critical this planning is in enterprise deployments.

    Test under realistic load before go-live. Not twenty items. The actual volume you expect in week three after rollout.

    Frequently Asked Questions

    What are Power Automate throttling limits and why do they cause flows to fail?

    Power Automate throttling limits are restrictions on how many actions or API calls your flow can make within a given time window. There are two layers: platform-level limits set by Microsoft based on your licence tier, and connector-level limits imposed by external services.

  • Agentic Workflows Are Not Just Fancy Automation

    Agentic Workflows Are Not Just Fancy Automation

    The mistake I keep seeing

    A client comes in. They’ve heard about AI agents. They want to ‘add AI’ to their approval workflow. So they take the existing 10-step Power Automate flow, stick a Copilot Studio agent somewhere in the middle, and call it an agentic workflow.

    It isn’t. It’s just the old process with a chatbot attached.

    This is the most common mistake I see right now, and it’s costing teams time and credibility. The agent becomes a fancy input form. The process stays broken. And when it fails — and it does — everyone blames the AI.

    What actually makes a workflow agentic

    An agentic workflow is not about adding a language model to a flow. It’s about giving the system the ability to reason about what to do next, not just execute a predefined sequence.

    The difference matters. In a traditional flow, you define every branch. Every condition. Every outcome. The machine follows instructions. In an agentic workflow, the agent interprets a goal, decides which tools or actions to use, and adjusts based on what it gets back.

    That requires a fundamentally different design approach. You’re not mapping steps — you’re defining boundaries, tools, and acceptable outcomes.

    Three things that have to change in your process design

    • Stop thinking in sequences. Agentic workflows are goal-driven, not step-driven. Define what done looks like, not every micro-step to get there. If your flow diagram looks like a subway map, you’re still in traditional automation mode.
    • Give the agent real tools, not just data. An agent that can only read a SharePoint list and send an email is not doing much reasoning. It needs to call APIs, query systems, write back to records, trigger sub-flows. Tool design is where most implementations fall apart — people give agents access to everything or nothing. Neither works.
    • Build in failure handling at the goal level. Traditional flows handle errors at the step level — if this action fails, go here. Agentic workflows need you to think about what happens when the agent reaches a dead end, produces a low-confidence result, or loops without resolution. I’ve seen agents spin for 40 iterations on a task that should have escalated to a human after three.

    Where this actually works in business processes

    Not everywhere. I want to be direct about that.

    Agentic design makes sense when the process has variability that you cannot fully predict upfront. Invoice exceptions. Complex customer complaints. Multi-system data reconciliation where the right answer depends on context you only know at runtime.

    It does not make sense for processes that are well-defined and stable. If your purchase order approval follows the same 6 steps every time, a standard Power Automate flow is the right tool. Don’t add an agent to it just because you can.

    The teams that get the most out of agentic workflows are the ones who identify a process where exceptions are eating their staff’s time, then let the agent handle the exceptions rather than replacing the whole flow.

    The orchestration layer nobody talks about

    When you start running multiple agents — one for document processing, one for customer communication, one for system updates — you need something coordinating them. This is where I see projects go sideways fast.

    In Copilot Studio and Power Platform, you can build orchestrating agents that hand off to specialist agents. But the handoff logic, context passing, and failure recovery across agents is not something the platform handles automatically. You have to design it. Most tutorials skip this. Then your multi-agent setup breaks in production because Agent B has no idea what Agent A already tried.

    Document your agent boundaries explicitly. What does each agent know? What can it do? What should it never do? Treat it like designing a team of junior staff who are fast and tireless but have no common sense unless you’ve given them the right context.

    Start smaller than you think you should

    Pick one process. One that has clear exceptions, high manual effort, and a measurable outcome. Build the agent, give it two or three tools, test it against real historical cases before you deploy it anywhere near live data.

    The teams that succeed with agentic workflows in 2025 are not the ones with the biggest ambitions. They’re the ones who are rigorous about scope, honest about where the agent is making decisions versus guessing, and fast to pull the agent out of the loop when something looks wrong.

    Agentic is a design philosophy. Apply it where it earns its complexity.