Pathfinder Blog

 Catch the recording Previously, on Dev Meetups…We did a full walkthrough of a meeting summariser AI Agent. We demoed how Mozart Orchestrator is able to orchestrate the entire process, and how to evaluate your AI Skill prompts in Automation Anywhere.🏗️ The Use Case: MeetingMind Task ManagementAt our hypothetical company, MeetingMind, decisions get made in meetings, mentioned in Slack threads, or buried in Zoom transcripts, and none of it makes it into the system that's supposed to track it. A PM ends up manually reviewing everything and updating tasks, risks, and action items after every sync.The goal: build an agentic process that takes meeting transcripts and channel messages as inputs, reasons over them, and writes structured updates — new tasks, flagged risks, assigned action items — back into MeetingMind tracking system, with one rule: if the transcript and the channel messages contradict each other, stop and surface the conflict to a human.It's a good candidate for an agent spec

If there’s one thing I’ve learned from leading community programs, it’s that the most exciting moments usually happen when you give brilliant people a tough problem, hand them the right tools, and get out of their way.That was our approach for at our first-ever in-person Pathfinder Hackathon in Bengaluru. Good news: our community delivered!Over two intense days on July 4 and July 5, developers, architects, and automation engineers came together to move past sandbox experiments and build enterprise-grade agentic solutions using the Automation Anywhere platform.Testing an AI agent in a controlled sandbox is easy but building one that handles real operational chaos, like supply chain surprises, event layout shifts, or sophisticated hiring fraud is an entirely different challenge. The Hackathon challenges While participants knew they’d be building working agentic solutions, we kept the exact problem statements hidden until the kick-off on Day 1. We revealed these three real-world operation

Catch the recording at .Previously, on Dev Meetups…At our June Dev Meetup, Charlie Needelman (Principal Product Manager, AI Agents), walked us through a credit hold resolution workflow, and how AI Agents can use Claude. The Scenario A customer bought a large MRI system, got flagged for exceeding their credit limit, and it turned out the invoice had incorrectly marked them as taxable — when they're actually a non-profit with a valid exemption certificate. The discrepancy came in through different formats from different systems: a Salesforce record, SAP invoice data, a SharePoint PDF, a ServiceNow support ticket.If your inputs are consistent, your format is predictable, and your conditions are manageable, a deterministic workflow can handle it. But AI Agents earn their place when things get variable. When a support ticket says "exempt" and an email says "not taxable" and a PDF says something else entirely, a hard-coded lookup fails. An Agent handles it because it can decode the meaning f

Previously, on Dev Meetups...https://youtu.be/6BisIdaZbVwOur community member, Javier Alverson, asked Sr Dev Evangelist, ​@Max.Cassidy an important question:if you didn't work at Automation Anywhere, what would you build with Mozart Orchestrator?What did Max say, I hear you ask.Two words. Grocery. Shopping. 🛍️🚫In his own words, Max Cassidy hates the grocery store, gets distracted by packaging, and buys too many sweets.So he built a workflow that will ease his life: one Excel file, an execute button, AI skills that find substitutions for out-of-stock items and surface the best prices, and a Process built on Mozart Orchestrator to orchestrate them all.And while it is debatable that there is such a thing as too many sweets, I think we’ll all agree that his agentic workflow is, at its core, a genius inventory management workflow. And the same event-driven logic, parallel branches, and AI-assisted decision-making can be applied to your enterprise workflows. Read on to know how. The Setup:

If you are managing an automation pipeline, your intake queue is likely overflowing with requests from business leaders who want to suddenly "add AI" to their departments. But your next agentic process automation use case is probably already sitting in your existing backlog.When new technologies emerge, Center of Excellence (CoE) teams and citizen developers often feel pressure to invent entirely new workflows to prove the technology works. In practice, the best opportunities for agentic process automation are the projects you already know about, specifically the ones that were simply too messy for standard rules-based automations to handle.I have seen firsthand that finding your first agentic use case requires a structured audit of your past rejections, your present bottlenecks, and your future intake pipeline. The Past: Mining the GraveyardRobotic process automation (RPA) is a powerhouse for executing highly structured, repetitive tasks at incredible scale, but it inherently requires

If you missed this month's Product Club, here's the honest summary: Nishikanth N, Director of Product Management, took us inside Mozart Orchestrator and showed us five new capabilities that just landed in version v.40, plus a sneak peek at something even more exciting coming in 41. Matt Stewart was also on hand to drop some real-world context that made the whole thing land differently.Read through what we covered, or watch the recording:Why orchestration even mattersNishi opened with a scenario most of us have lived: a customer complaint about a wrong item delivered. With traditional task automation, the process falls apart the moment it needs to wait on something, like a supplier response that takes two days, or a routing rule buried so deep in bot code that changing it requires a full redeploy.Matt added to this from his own experience managing an RPA program: bots would do what they could, dump the rest into an exceptions file, and hand it to humans to finish manually. "People's job

Hey Pathfinders!This month's Product Club brought us something a little different. With my absence, developer evangelist Max stepped in as host, joined by Rinku Sarkar, Director of Product Management, to walk us through two big topics: AI Evaluations and the upcoming AI Agent Logs. Buckle up, because this one was packed. For the full experience, watch the video below:The core tension for enterprise AIRinku opened with a number that's worth sitting with: research found that even top AI models only succeed on structured output tasks about 75% of the time. That's one in four attempts going sideways. In a process touching financial data, customer workflows, or compliance? That's not a rounding error, that's a real business risk.And here's the thing: this isn't a vendor problem. It's not a maturity problem. It's just how AI works. AI is probabilistic. It's making an educated guess every single time, which is exactly what makes it powerful and exactly what makes it hard to trust at scale.Rin

Hey Pathfinders!If you missed May's Product Club, don't worry, we've got you covered. Hosted by Shreya and joined by Arun K. Mudaliar, Principal Product Manager at Automation Anywhere, this month's session was all about COE Manager and some shiny new features that just dropped in version 40. Here's what went down:(for the full experience, watch the recording below) CoE Manager provides you with visibility into your programAs automation programs grow from a handful of bots to dozens (or hundreds!), the question shifts from "can we build more?" to "how do we know what we've built is actually working?" Governance gets messy, ROI gets murky, and suddenly your pipeline is running on gut feeling instead of data. That's exactly the gap COE Manager was built to close.CoE Manager tackles your governance and pipeline issuesCOE Manager tackles two big challenges: pipeline and governance (managing your entire automation estate from idea to deployment) and tracking KPIs and realizing value (going b

We conducted an outstanding Pathfinder Meetup in Oregon, co-hosted at Nike's campus in Beaverton, bringing together automation and AI practitioners for of real-world use case sharing. The event featured two MVPs, a Technical Advisory Council member, customers, partners, leaders, and practitioners. With perspectives spanning production agentic workflows, finance automation, healthcare intake transformation, GenAI-driven collections, and a forward-looking framework for how organizations to progress from RPA to agentic processes, it was a highly productive day.Micah Smith, Automation Anywhere's VP of Developer Relations and Community set the stage with his presentation on agentic use case identification and the progression from RPA to Process to Agentic Process giving attendees a clear framework for thinking about where their automation journeys are headed and how to identify the right opportunities for agentic transformation.TAC member Matt Manning and his Nike colleagues Sean Lewis and

“How to Be Your Own Advocate in the Workplace”What an incredible session. The energy, honesty, and vulnerability in the room reminded us exactly why this community exists. Here's a recap of the key moments and takeaways of the session.Why this topic?During our Pathfinder Summit survey, we asked: Have you ever asked for a raise? 84% of respondents said no. That number sparked this conversation and as we dug deeper in our prep sessions, we realized it's about so much more than salary. It's about knowing your worth, advocating for yourself, and asking for what truly motivates you.Danielle Deriso: “Don't Accept No as the Final Answer”Danielle, Senior Manager of Finance Digital Governance and MVP, shared how she turned an organizational change into an opportunity to champion AI at her company even after being told the role was "two years out."Her key takeaways:See something, say something. See a gap, fill a gap. Don't wait for permission to move in the direction of your goals. A "no" is oft

If you run an automation Center of Excellence (CoE), you are likely intimately familiar with the end-of-quarter ROI spreadsheet. You spend hours compiling the number of manual clicks avoided and multiplying them by an average hourly rate to prove your program's worth to a skeptical executive team. For over a decade, that simple calculation has been the gold standard.Robotic process automation (RPA) is incredibly effective at executing structured, high-volume tasks, and measuring those executions in "hours saved" makes perfect sense. However, agentic process automation (APA) expands the aperture of recognizable benefits. To accurately measure the return on investment (ROI) for AI agents, automation programs must look past baseline efficiency and adopt a multi-dimensional scorecard.By categorizing impact across the three pillars of the Agentic Impact Measurement (AIM) Framework, CoE leaders can communicate the true strategic value of their deployments. Pillar 1: Financial and Operational

There's a moment in every technology shift when the people who get there first stop being early adopters and start being the ones everyone else calls.We think that moment for AI automation is now. And we're building the credential to mark it.At Imagine Dallas and Imagine India, we're launching the AI Automation Engineer Certification. The people who complete it will be the first certified Automation Anywhere AI Automation Engineers in the world. What Is the AI Automation Engineer Certification?The AI Automation Engineer Certification is designed for the people want to lead AI adoption, architect AI projects, and build the agentic automations that will accelerate results for their organizations. A certified AI Automation Engineer can do three things well: identify the right opportunities for agentic automation and make the business case for them, design and build the solution, and evaluate whether it's actually working. That combination — strategic judgment and technical execution — is

When you’ve been immersed in new ideas, whether you're spent days at a conference, finished a deep-dive YouTube rabbit hole, or walked away from a conversation that finally clicked, inspiration makes your brain feel excited about the potential of what’s next. But it’s hard to get started with what’s next when you still have daily stand-ups, full email inboxes, and a backlog that didn't pause while you were thinking big.I know this feeling, and I also know that making progress requires a structured plan to move ideas into action. For those who attended Imagine, I wanted to make sure that they left Dallas understanding the power and potential of AI agents, and how to identify also which processes actually require an AI Agent. So, that’s what I spoke about.I believe that if you can learn to identify the symptoms of an agentic process, you can accurately diagnose opportunities and prescribe exactly where or whether an AI agent belongs in a given workflow.The diagnostic isn't complicated, b

Automation 360 provides templates that you can use to quickly create and build new automations. Templates contain a prebuilt automation code structure as per the specific automation use case. To learn more, click here Automation Templates. Soon, Automation Anywhere will provide predefined templates for common automation use cases. These templates will offer developers a starting point and a sample to explore while building their automations. Would you like all developers to have access to these predefined templates by default, or should the CoE review them before being made available? Vote below and share any relevant details & considerations by replying to this thread. 

If you are managing an automation Center of Excellence, you are likely feeling intense operational pressure to scale AI agents across your business units without breaking your existing tech stack. It is easy to build a single standalone pilot, but orchestrating multiple intelligent systems across legacy environments while maintaining security creates immediate development bottlenecks.At Imagine 2026 in Dallas, our Chief AI and Development Officer Adi Kuruganti detailed the exact architecture making this possible. The system relies on three core pillars: universal orchestration, contextual intelligence, and centralized governance.These pillars function like the three legs of a stool. Each one is essential on its own, but working in tandem, they facilitate the vision of agentic process automation (APA). This unified foundation is exactly how organizations transition to the autonomous enterprise—an AI-first model where up to 80 percent of work runs autonomously or is AI-assisted with a hu

In our recent user group session, our Community Captain and Agentic Bounty Challenge winner Akbar Shaikh, RPA Principal Architect at Novatio Solutions, showcased one of the agents he built — a live demo of his Customer Complaint Routing Agent.Akbar's client was receiving hundreds of survey responses daily with no efficient way to triage them. Teams spent hours manually reading free-text feedback, assessing severity, and routing complaints — often too late to make a difference. The goal was simple: have an AI agent read the feedback, classify how serious it is, and alert the right person immediately. What He BuiltThe agent pulls customer feedback from SurveyMonkey via API and classifies each response into one of three tiers: normal, complaint, or critical. Normal covers positive feedback and minor suggestions; complaint flags clear dissatisfaction; critical catches aggressive language or sentiment that demands an immediate response.Four tools power the workflow: a Survey Tool to retriev

Catch the recording at https://youtu.be/39LMEkEpQLw?si=qeRM1ZHkGO2af9HqPreviously, on Dev Meetups...We did things differently this time around. By now, most of you are familiar with AI Agent Studio and what it can do. So this time, we handed the agent a full toolkit and no instructions on how to use it, let the community vote on the design decisions, and watched it figure out the rest.The SetupThe session was built around AI Agent Studio and a fictional retail dataset: Max's Bargain Homewares, a fully relational customer support application with order history, a product catalog, complaint records, and a work queue. Matt had pre-built a set of API task tools wired to this backend. The community helped us decide what kind of agent to build with them.Before getting to the polls, we walked through something worth sitting with: the difference between automating with a legacy mindset versus an agentic one. The legacy approach scopes by team — build the customer investigations process, then p

Challenge #3 - Automation Resiliency  We're excited to present the third and final challenge of Bot Games Season 4. Our aim in this challenge is to showcase how a Generative AI model can enhance automation resilience. When target applications are updated, the object paths for elements may change, requiring updates to the corresponding automations. By utilizing a Generative AI provider like ChatGPT (or any other provider of your choosing), we can input the source code of a web page and request object paths for all the necessary elements. These paths can then be used to automate interactions with the web page.OpenAI Generative AI package is available in Bot Store and you can read the included documentation or check out our quickstart video showing the full setup process. If you are using your own enterprise version of Control Room, you can download the OpenAI Generative AI package from the Bot Store and import it. If you are using the Community edition, it is already pre-loaded. Challeng

There's a narrative forming in the market that agentic process automation (APA) requires a completely different talent profile than what most automation teams have today.That narrative misses the point that people who have been building and running robotic process automation (RPA) programs for years already carry many of the instincts that agentic systems demand.Process thinking, exception handling, understanding where decisions occur within a workflow, and knowing when something needs human judgment and when it can be automated. These are foundational to how agents operate, and they're skills that take years to develop through real-world practice.You already have a headstart, but you might be wondering how to build on your foundation today.Process fluency is the hardest skill to hire forIn this episode of the Agentic Edge Podcast, Mike Reynolds, a business and technology executive with over 20 years of experience in financial services at KeyBank, shared something that reinforced a pat