Genie X Hub represents Project : Eclipse Hackathon

Genie X Hub Presents: PROJECT ECLIPSE HACKATHON

Build AI That Thinks, Acts & Automates — Not Just Answers.

Project Eclipse Hackathon is a hybrid AI-focused hackathon by Genie Hive Private Limited, designed for college students, developers, AI enthusiasts, designers, and innovators who want to build practical solutions using Generative AI, Agentic AI, Autonomous Systems, and Intelligent Automation.

This is more than a conventional chatbot-building competition. Participants are challenged to build systems that can understand problems, reason, use tools and APIs, execute actions, verify outcomes, and involve humans when necessary.

The hackathon features 5 curated real-world AI challenges + 1 Open Innovation challenge.

Whatsapp Community: Join this community for all the event updates  : https://chat.whatsapp.com/CwOLKlgym1W4TpcNDx8TUC

THE ECLIPSE CHALLENGE

AI is moving beyond systems that simply generate responses.

Project Eclipse focuses on systems that follow the workflow:

INPUT → UNDERSTAND → PLAN → USE TOOLS → ACT → VERIFY → RECOVER / ESCALATE

Participants are encouraged to build working AI systems that solve meaningful problems and demonstrate real task completion rather than simply generating convincing text.

CHALLENGE TRACKS

Participants can choose one of five curated challenges or participate in the OPEN-X Open Innovation Challenge.

01 — AI Agent for Production Incident Investigation & Recovery

Build an autonomous production-operations agent that investigates a simulated production incident across application logs, infrastructure metrics, deployment history, Git commits, alerts, and service dependencies.

The system should demonstrate capabilities such as:

• Multi-source investigation
• Tool/API calling
• Reasoning over temporal events
• Root-cause hypothesis generation
• Confidence estimation
• Remediation planning
• Human approval for high-impact actions
• Recovery verification
• Post-incident report generation

02 — Autonomous Resolution of Cross-System Transaction Mismatches

Build an AI agent that investigates transaction mismatches across invoices, payment records, purchase orders, emails, spreadsheets, and accounting/ERP systems.

The system should be able to:

• Ingest data from multiple sources
• Match transactions/entities
• Retrieve supporting evidence
• Classify discrepancies using context
• Detect partial, duplicate, or missing payments
• Recommend appropriate actions
• Maintain an auditable action trail
• Use human approval for consequential updates

03 — From Contract Language to Executable Business Actions

Build an AI agent that converts contracts into executable compliance workflows.

Your system should be able to:

• Extract obligations
• Identify responsible parties and deadlines
• Reason over conditional clauses
• Generate tasks and workflows
• Send reminders and escalations
• Maintain compliance records
• Handle document hierarchy and amendments
• Resolve conflicts between contract documents and instructions

04 — Autonomous Requirement-to-Quotation Procurement System

Build an AI procurement agent that takes a natural-language purchase requirement and analyzes vendor quotations.

The system should:

• Understand requirements
• Parse quotations from documents, emails, or structured files
• Normalize specifications
• Check mandatory constraints
• Compare cost and trade-offs
• Detect anomalies
• Simulate vendor clarification/negotiation
• Prepare a recommendation for human approval

The goal is not simply to identify the cheapest quotation, but to determine whether a quotation actually satisfies the required specifications.

05 — Issue-to-Pull-Request AI Engineering Agent

Build an AI software-engineering agent that receives a repository, issue description, and acceptance criteria and takes the problem from:

ISSUE → INVESTIGATION → IMPLEMENTATION → TESTING → PULL REQUEST

The agent should demonstrate:

• Repository exploration
• Issue and acceptance-criteria understanding
• Technical hypothesis formation
• Code modification
• Test execution
• Failure analysis and iteration
• Patch/PR generation
• Human review before deployment

OPEN-X — OPEN INNOVATION

Have your own problem worth solving?

Build beyond the brief.

In OPEN-X, participants can define their own real-world problem and build an original solution using Generative AI, Agentic AI, Autonomous Workflows, or Intelligent Automation.

Your submission should explain:

Problem — What real-world problem are you solving?

Target Users — Who experiences this problem?

Existing Gap — Why are existing approaches insufficient?

Proposed Solution — What are you building and how does it work?

AI / Agent Component — Why does the solution meaningfully require GenAI, agents, or intelligent automation?

Demonstration — Can you demonstrate the system working end-to-end?

Impact — What changes if your solution is deployed?

A conventional application with an AI label added on top is not sufficient for the Open Innovation challenge.

️ HACKATHON TIMELINE

Registration

20 September – 30 September 2026

Online

Round 1 — Idea Submission

01 October – 06 October 2026

Online

Shortlisted Teams Announcement

09 October 2026

Round 2 — Building Phase

11 October – 23 October 2026

Online

Finalists Announcement

25 October 2026

Grand Finale

27 October 2026

Offline — Chennai, Tamil Nadu, India

Grand Finale venue: To Be Announced

All preliminary rounds will be conducted online. The Grand Finale will be conducted offline in Chennai.

Whatsapp Community: Join this community for all the event updates  : https://chat.whatsapp.com/CwOLKlgym1W4TpcNDx8TUC

ROUND 2 PARTICIPATION FEE

₹499 INR

Initial registration and Round 1 participation are FREE.

Only teams that are shortlisted for Round 2 will be required to pay the ₹499 INR Round 2 participation fee to enter the building phase.

The fee is not applicable during initial registration or Round 1.

Participation flow:

Register → FREE

Round 1 Idea Submission → FREE

Shortlisted for Round 2 → ₹499

Round 2 Building Phase → Build & Submit

Selected as Finalist → Grand Finale

️ HOW THE HACKATHON WORKS

ROUND 1 — IDEA SUBMISSION

Participants select one of the five curated challenges or choose the OPEN-X track.

Teams submit their proposed solution, explaining the problem, approach, AI/agent architecture, workflow, and expected impact.

Submissions will be evaluated based on problem relevance, proposed approach, technical potential, use of AI/agentic systems, and overall solution quality.

Shortlisted teams will advance to Round 2.

ROUND 2 — BUILD

Shortlisted teams will transform their ideas into a functional and demonstrable prototype.

Participants should aim to demonstrate an end-to-end system rather than a concept or basic chatbot.

For agentic solutions, the recommended execution flow is:

INPUT → PLAN → TOOL USE → DECISION → ACTION → VERIFICATION

Systems should demonstrate meaningful task completion, reliability, and appropriate handling of failures or situations requiring human intervention.

GRAND FINALE

Finalists will present their working solutions to the judging panel at the offline Grand Finale in Chennai, Tamil Nadu.

Final presentations will focus on:

• Working product demonstration
• Technical implementation
• Agentic reasoning
• Tool/API integration
• Autonomous execution
• Reliability and failure handling
• Innovation
• User experience
• Real-world applicability

JUDGING CRITERIA

Criterion

Weight

Problem Understanding & Relevance

10%

Agentic Reasoning & Planning

20%

Tool / API / System Integration

15%

Autonomous Task Completion

20%

Reliability & Failure Handling

15%

Innovation

10%

UX / Demonstration

10%

Total

100%

These criteria are based on the official challenge framework for the hackathon.

SUBMISSION EXPECTATIONS

Participants should prepare a working prototype or demonstrable implementation.

Depending on the selected challenge, submissions may include:

• Working prototype
• Problem and solution explanation
• Architecture/workflow diagram
• AI/agent architecture
• Tools and APIs used
• Repository/code submission where required
• Demo video or live demonstration
• Known limitations
• Safety controls
• Future scope

For agentic submissions, participants are encouraged to make the execution trace visible:

INPUT → PLAN → TOOL USE → DECISION → ACTION → VERIFICATION

WHO CAN PARTICIPATE?

Project Eclipse Hackathon is open to college and university students.

Participants interested in the following areas are encouraged to participate:

• Generative AI
• Agentic AI
• AI Automation
• Machine Learning
• Software Engineering
• Data & Analytics
• Developer Tools
• Enterprise Automation
• Product Development
• Emerging Technologies

You don't need to be an AI expert.

If you can identify a meaningful problem and build a solution around it, Project Eclipse is for you.

️ TECHNOLOGIES

Participants may use appropriate technologies, frameworks, APIs, cloud services, open-source models, datasets, and development tools, subject to the final hackathon rules and applicable usage limits.

Potential areas include:

LLMs • AI Agents • RAG • APIs • Automation • Computer Vision • NLP • Voice AI • Multi-Agent Systems • Generative AI • Cloud • Full-Stack Development • Developer Tools

PRIZE POOL

TOP 1

₹10,000 Cash Prize

Trophy

Certificate

TOP 2

₹6,000 Cash Prize

Trophy

Certificate

TOP 3

₹4,000 Cash Prize

Trophy

Certificate

TOP 4 & TOP 5

Goodies

Certificate

PARTICIPANTS

Certificate for all participants

FINALISTS

Certificate for all finalists

HACKATHON FORMAT

Hybrid Hackathon

Registration — Online

Round 1 — Online

Round 2 — Online

Grand Finale — Offline, Chennai, Tamil Nadu

OFFICIAL HACKATHON WEBSITE

https://eclipse-hackathon.vercel.app/

READY TO ENTER THE ECLIPSE?

The next generation of AI isn't just about generating answers.

It's about building systems that can:

REASON → ACT → AUTOMATE → VERIFY → DELIVER

Choose your challenge.

Build your solution.

Put it to work.

Welcome to Project Eclipse.

Whatsapp Community: Join this community for all the event updates  : https://chat.whatsapp.com/CwOLKlgym1W4TpcNDx8TUC