UiPath AgentHack is the UiPath Community global online hackathon for developers, automation engineers, data scientists, AI engineers, students, and professionals who want to go beyond basic automation and build agentic solutions that work in enterprise environments.

Over 7 weeks, you'll design and ship a real, working solution on the UiPath Platform™. Not a concept, nor a slide deck, but a working solution that handles complexity, survives interruptions, keeps humans in the loop, and solves something that matters. Uipath AgentHack challenges you to build, run, and orchestrate real agentic solutions on the UiPath Platform, the single enterprise control plane that coordinates AI agents, automations, people, and applications end to end. Whether you build with UiPath's native capabilities or combine them with frameworks like LangChain, CrewAI, or AutoGen, your solution must use UiPath as the execution and orchestration layer.

We have $50,000 in cash prizes and three challenge tracks built around the most critical capabilities in enterprise agentic AI.

Pick the track that best fits your project and submit:

• Track 1: UiPath Maestro Case
• Track 2: UiPath Maestro BPMN
• Track 3: UiPath Test Cloud

About the challenge

Coding agents have redefined how we build and the real value now lies in how we operate and govern agents at scale, bridging the gap between a prototype on a laptop and software running in production.

UiPath AgentHack is built around that challenge. At the center of this hackathon is UiPath for Coding Agents, a platform-wide capability that enables developers to use coding agents (Claude Code, Codex, Cursor, and Gemini CLI, etc) to build, test, deploy, operate, and govern enterprise automations. Using natural language, you'll design AI agents that drive tangible business value, with full support for external frameworks such as LangChain, CrewAI, and AutoGen.

The three challenge tracks are built around real-world enterprise problems that require coordination, adaptability, and orchestration at scale: structured process orchestration with Maestro BPMN, agentic case management for dynamic and exception-heavy work, and agentic software testing for AI-driven automations.

Throughout the 7-week hacking period, UiPath will host community meetups, office hours, and will offer enablement resources to help you get up to speed with the platform and refine your solution.

Track 1: UiPath Maestro Case. Build a solution that orchestrates dynamic, exception-heavy business processes using UiPath case management capabilities. Your solution should move work through stages, involve handoffs between agents, robots, and people, and keep humans in charge at key decision points. Agents within your case flow can be built on UiPath or an external framework; the platform handles coordination regardless of where the agents come from. 

Think about scenarios like: insurance claims processing where cases move through intake, investigation, and settlement stages; patient care coordination where agents manage referrals, scheduling, and follow-ups across providers; HR onboarding workflows where each new hire case progresses through document collection, system provisioning, and training assignment. 

Track 2: UiPath Maestro BPMN. Build a solution that models and runs an end-to-end business process using BPMN 2.0 in UiPath Maestro. Your process should orchestrate humans, robots, agents, and APIs through a defined flow with clear tasks, decisions, and handoffs. Agents within the process can be built with UiPath Agent Builder, coding agents, or external frameworks like LangChain, CrewAI, AutoGen, or any other agent platform. We want to see processes that move work cleanly from start to finish, with the right actor doing the right task at the right time. 

Think about scenarios like: An order-to-cash workflow where RPA pulls orders from email, EDI, and portals, an agent normalizes line items and flags pricing exceptions, and Maestro BPMN orchestrates credit checks, inventory allocation, fulfillment, and invoicing across ERP and CRM in a defined sequence. A collections agent monitors aging invoices and escalates disputes to finance for review. Or a procure-to-pay process where agents parse requisition intent, recommend vendors, and route approvals based on budget thresholds and category rules, RPA handles PO creation and invoice ingestion across ERP and AP systems, and an invoice agent reconciles discrepancies between PO, receipt, and invoice, escalating only true exceptions to AP for human review. Maestro BPMN keeps the full flow coordinated from requisition to payment

Track 3: UiPath Test Cloud. Create agents that use UiPath Test Cloud to reimagine how software testing is designed, automated, executed, and managed across modern enterprise environments. 

Your goal is to show how agentic software testing can improve quality across AI-driven automations, enterprise applications, and connected business workflows. These agents should help you move faster with more confidence by increasing coverage, improving reliability, and reducing the manual effort required to validate complex systems.

Think of building agents that can: evaluate requirements and turn them into meaningful test scenarios, identify fragile or outdated tests before they slow down a release, recommend fixes when automation breaks, or help orchestrate the right tests at the right time based on risk, coverage, and change impact. You might also explore how agents can validate AI-infused workflows, including third-party agents or AI services that participate in a UiPath-orchestrated process.

All solutions must run on the UiPath Automation Cloud. You can include Agent Builder, Maestro, API Workflows, coding agents, and RPA where needed. You're welcome to bring in agents built on external frameworks and LLMs, in fact, we encourage it. The point is that UiPath is the orchestration and governance layer that ties everything together. 

Tip: If your process has unpredictable paths that emerge as the work unfolds, choose Track 1 - UiPath Maestro Case. If your process has a predictable sequence you can map in advance, choose Track 2 - UiPath Maestro BPMN.

Bonus: solutions that use coding agents through UiPath for Coding Agents (Claude Code, Codex, Cursor, Gemini CLI, etc) will receive additional points during judging. We're especially interested in seeing how participants combine coding agents with low-code components, or blend UiPath-native agents with external agents, to solve complex problems

You can't submit the same project to multiple tracks, and your solution needs to clearly align with whichever track you choose, both in how it's built and how it's described.

Get started

• Register on Devpost for UiPath AgentHack.
• Form or join a team of 1 to 4 people and pick your track.
• Ask for access on UiPath Labs here: Teams must designate a representative to complete the access form on behalf of the group. Within 3 business days of asking for access, you'll receive a separate email with your UiPath Labs access link and credentials. UiPath Labs come fully equipped with agentic and AI units, so you can build without constraints.

• Business Impact & Adoption Potential
  Judges will assess the real-world relevance and production viability of the solution as presented, including the clarity of the business case and the scalability of the approach.
• Platform Usage
  Judges will assess the depth and deliberateness of UiPath platform usage as demonstrated in the live presentation, including the use of Agent Builder, Maestro, API Workflows, coded agents, and external frameworks where applicable.
• Technical Execution, Feasibility & Versatility
  Judges will assess architectural soundness, code quality, and production-readiness as demonstrated through the live demo, including the solution's handling of exceptions, failures, and edge cases.
• Completeness of Delivery
  Evaluators will assess whether the submission includes a functional end-to-end prototype, a public GitHub repository with a README and setup instructions, and a demo video of no more than five minutes.
• Creativity & Innovation
  Judges will assess the originality of the approach as demonstrated in the live presentation, including novel design decisions, unexpected orchestration patterns, and creative problem framing.
• Presentation
  Judges will assess the clarity, structure, and delivery of the live presentation, including the logical flow from problem to solution to impact, the confidence and coherence of the demo, and the accessibility of technical communication.
• BONUS POINTS: CODING AGENTS
  Submissions that demonstrate the use of coding agents through UiPath for Coding Agents, including Claude Code, Codex, Cursor, and Gemini CLI, will receive additional points within the Platform Usage criterion in both Phase 1 and Phase 2 judging.