展开原文（5685 字）
The Great Agent Hackathon, presented by TGPF 2026
The Great Agent Hackathon is a 24-hour, in-person build happening overnight in Bangalore, as part of The Great Product Festival, India's largest product and startup conference.
Teams get real problem statements, real mentorship through the night, and a live Main Stage finale in front of India's top product leaders at TGPF.
This year's hackathon is powered by
Freshworks
as title sponsor, with
ElevenLabs
powering the build.
Problem Statement Tracks
• Track 1 — Customer & Employee Experience: agents built on Freshworks Agent Studio, MCP, and multi-agent orchestration.
• Track 2 — Platform Agent Skills & Knowledge: reusable skills, MCP integrations, and the Freshworks developer platform.
• Track 3 — AI-native Enterprise (Open): open canvas, any stack, greenfield agentic products.
Examples
-
Track 1: Customer Experience & Employee Onboarding Agents
Voice Support Agent — handles a support conversation, identifies intent, takes actions and hands off to a human when needed.
Employee Policy Agent — answers HR/policy questions and proactively guides employees through relevant processes.
Track 2: Platform Agent Skills & Knowledge
Universal Company Knowledge Agent — connects Slack, docs, CRM, tickets and internal wikis to answer questions with context.
Knowledge-to-Action Agent — doesn't just retrieve information; converts company knowledge into actions and workflows.
Agent Handoff Skill — intelligently decides when one specialist agent should hand a task to another.
Track 3: AI-Native Enterprise Agents
AI Chief of Staff — monitors meetings, emails, tasks and priorities, then proactively drives follow-ups and decisions.
AI Operations Manager — monitors business metrics, detects anomalies and coordinates agents/teams to resolve them.
Autonomous Sales Agent — researches prospects, identifies intent, personalises outreach and manages follow-ups.
AI Finance Controller — monitors transactions, identifies anomalies, reconciles data and flags exceptions for human approval.
Requirements
Stage 1: Online Qualification Round
To qualify for The Great Agent Hackathon, teams must submit the following on Devpost before the qualification deadline:
A Figma prototype or working demo
of your idea addressing one of the 3 problem statement tracks. This does not need to be a finished product, a clickable prototype, wireframe flow, or early build is enough to show the idea clearly.
A short written submission
covering:
The problem statement your team is addressing
Your proposed solution and how it works
Why your team is the right one to build it
A brief video walkthrough
(1-2 minutes) explaining your prototype and idea. Loom, YouTube, or Vimeo links are fine.
Team details
: names, roles, and one line on relevant experience for each team member.
Submissions will be reviewed by the TGPF and sponsor team. Shortlisted teams will be notified and invited to Bangalore for the in-person build.
Stage 2: In-Person Build (Bangalore, Shortlisted Teams Only)
Shortlisted teams travel to Bangalore for the 24-hour overnight build as part of TGPF 2026.
At this stage, teams must submit:
A working prototype or demo
of the product built during the 24-hour build window, functional and ready to demo live.
An updated project description
covering what was built, the tech stack used, and how it evolved from the original Stage 1 idea.
Public code repository link
(GitHub or equivalent) with a clear README.
A live demo
on the Main Stage in front of judges and the TGPF audience.
Hackathon Sponsors
Prizes
₹
100,000
+ in prizes
+ other prizes
Overall Winner
₹
40,000
in cash
1 winner
Cash + ₹30,000 in ElevenLabs credits + ₹4,70,000 in Dodo Payments credits
Track 1 Winner
₹
20,000
in cash
1 winner
Cash + ₹ 2,30,000 Dodo payment credits
Track 2 Winner
₹
20,000
in cash
1 winner
Cash + ₹ 2,30,000 Dodo payment credits
Track 3 Winner
₹
20,000
in cash
1 winner
Cash + ₹ 2,30,000 Dodo payment credits
Best Project on Eleven Labs
1 winner
₹80,000 in credits
This price can be combined with Track winner prize
Top 20 teams
20 winners
₹ 47,000 Dodo payment credits each
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Minakshi Khuntia
Senior Director of Product Management - Freshworks
Amrit Raj
Founder - Future Minds Institute
Judging Criteria
Innovation & Originality
How creative and original is the idea? Does it approach the problem statement in a way we haven't seen before, or push beyond an obvious solution?
Technical Execution
How well is the project built? Judges will look at functionality, code quality, and whether the team pulled off what they set out to build within the time given.
Use of AI / Agentic Design
Does the project meaningfully use AI or agent-based workflows to solve the problem, rather than bolting AI on as an afterthought?
Relevance to Problem Statement
How directly does the solution address the chosen track, Tech in Product, Growth and Monetisation, Sponsor-Driven Challenge, or Open Track, and how well does it solve a real, clearly defined problem?
Presentation & Demo Quality
How clearly does the team communicate what they built, why it matters, and how it works during the live demo? Judges will weigh clarity and storytelling, not just polish.
Potential Impact
If this were taken further, how much real-world value could it create for the people or businesses it's meant to serve?
Questions?
Email the hackathon manager
Invite others to compete
Hackathon sponsors
This site is protected by reCAPTCHA and the Google
Privacy Policy
and
Terms of Service
apply.