Meet the Winners of the Agentic Bounty Challenge

At Automation Anywhere Pathfinder Summit 2025, we invited the Pathfinder Community to take on an exciting challenge: Identify a real business problem. Design a goal-based AI agent. Build something that works.

The response exceeded my expectations.

What impressed me most wasn’t only the technical creativity. It was the storytelling. Each participant clearly articulated the problem, explained why it mattered, and showed how an agent could step in to help. That mindset is what agentic process automation (APA) is all about.

Today, I’m excited to celebrate our three winners.
 

🥉 Third Place: Jenna – Turning Product Reviews Into Insight

Jenna (​@Jenna4321) started with a simple observation: every company collects product reviews. Very few teams actually extract structured insight from them fast enough to influence decisions.

Anyone who has tried to manually review hundreds of comments knows the drill. You read. You tag. You summarize. Hours disappear. The data sits there waiting.

So she built an agent to change that.

Jenna’s Product Review Data Insights Agent:

• Accepts CSV or JSON review data
• Performs automated sentiment analysis
• Adds a structured sentiment column to the dataset
• Identifies the top three most reviewed products
• Produces a clean, downloadable output ready for action

Instead of spending days analyzing text, teams get categorized insights in minutes.

What stands out here is the practicality. It’s a straightforward, business-ready agent that marketing, product, and operations teams could start using immediately.

👉 Download Jenna’s submission here.
 

🥈 Second Place: Omkar – A Metro System That Thinks Ahead

Omkar (​@OmkarMahajan) brought us into an environment many of us don’t think about daily: metro transit operations.

His story begins in a high-pressure setting. A staff member reports an incident. Time matters. Safety matters. Documentation matters. Multiple manuals exist. Decisions must be accurate and traceable.

That real-world complexity shaped his agent, Aura Metro.

When an incident is reported, the agent:

• Reviews both a business manual and a disaster management manual.
• Identifies recommended next steps.
• Creates tasks and work orders.
• Reviews historical part orders before suggesting a new purchase.
• Logs actions in Salesforce for traceability.
• Routes decisions through human approval.
• Produces a final execution summary.

One detail stood out during judging: before placing a purchase order, the agent checks order history and surfaces how often that part has already been requested. That layer of context helps teams avoid unnecessary spending and make smarter decisions.

This solution demonstrates thoughtful orchestration across systems, knowledge sources, and workflows. It also reinforces responsible design with human-in-the-loop approval and complete audit logging.

👉 Download Omkar’s submission here.
 

🥇 First Place: Ganesh – From Sprint Chaos to Structured Execution

Ganesh’s (​@Ganesh Bhat) story will feel familiar to anyone in Agile development. Picture a sprint planning session:

• Whiteboards covered in ideas.
• Sticky notes everywhere.
• Meeting transcripts.
• Follow-up emails.

Then comes the tedious part: someone must translate all of that into structured Jira epics, stories, and tasks.
That friction became his starting point for creating the Sprint to Jira AI Agent.

Ganesh’s Sprint to Jira AI Agent ingests:

• Meeting transcripts
• Unstructured notes
• Screenshots
• Emails

From there, the agent:

• Extracts structured requirements
• Organizes them into epics, stories, and tasks
• Aligns outputs to Agile best practices
• Includes human-in-the-loop approval
• Creates issues securely via API integration

What makes this solution compelling is its relatability. Nearly every software team faces this challenge. Ganesh translated a common pain point into a structured, agent-driven workflow that preserves governance while saving time.

👉 Download Ganesh’s submission here.
 

Explore the Submissions Yourself

All of the solutions highlighted here—along with many other Agentic Bounty Challenge entries—are available for free in the Automation Anywhere Agentic App Store.

The Agentic Bounty Challenge proved when you combine creativity with APA, real-world problems start looking much more solvable.

I can’t wait to see what you will build next.

P.S. You can check out the winner announcement video on LinkedIn – the reactions are priceless!