24 Hours
Continuous build and demo cycle.
April 16 - 17, 2026 | 24-Hour Offline Build Sprint
Build Industry-Ready Agentic AI Systems
A professionally mentored hackathon where teams design, build, and demonstrate production-grade AI agents with measurable business outcomes.
Prize Pool: Rs.50,000
Internship Offers: 5 participants
Team Size: 2-3 members
Overview
A recruitment-first hackathon with real-world evaluation
HackStrom'26 is designed to assess practical AI engineering capability. Teams will create complete solutions under strict time constraints, present architecture decisions, and defend trade-offs before an overseas jury and industry mentors.
5 Tracks
Agentic AI and intelligent systems.
Rs.100
Registration fee per team.
2nd / 3rd Year
Eligibility for participation.
Problem Statements
Hackathon Tracks
Each track requires a working backend, interactive frontend, and transparent reasoning traces.
Problem Statement
Build an AI agent that dynamically selects optimal shipment routes (air, sea, road) based on constraints such as cost, service-level agreements (SLA), and external conditions.
Example Input
"Ship 500kg from Dubai to Rotterdam within 5 days under $4000"
Expected Capabilities
- Multi-step planning for route selection.
- Cost versus SLA trade-off analysis.
- Integration with external tools such as pricing APIs, weather data, and port congestion data.
- Structured output including route selection, cost breakdown, ETA, and reasoning.
Technical Requirements
- Backend API (FastAPI recommended, but not mandatory).
- Frontend application (React recommended, but not mandatory).
- Agent must evaluate multiple route options and justify its decision.
Problem Statement
Build an AI agent that predicts shipment delays and recommends proactive mitigation strategies.
Example Input
"Shipment arriving at Jebel Ali in 3 days - identify risks"
Expected Capabilities
- Analysis of historical shipment data.
- Integration with external signals such as weather and news.
- Prediction of delay probability and contributing risk factors.
- Recommendations for rerouting or mitigation.
Technical Requirements
- Backend API (FastAPI recommended, but not mandatory).
- Frontend dashboard (React recommended, but not mandatory).
- Agent must utilize at least two data sources and output a risk score with explanation.
Problem Statement
Build an AI agent that extracts, validates, and generates customs-compliant documentation.
Example Input
Upload invoice and bill of lading documents.
Expected Capabilities
- Data extraction from documents.
- HS code classification.
- Validation against country-specific compliance rules.
- Generation of structured customs declarations.
Technical Requirements
- Backend API with file upload support (FastAPI recommended, but not mandatory).
- Frontend interface (React recommended, but not mandatory).
- Integration of OCR/NLP pipelines and rule-based agentic validation.
Problem Statement
Build an AI-powered control tower system that monitors shipments and recommends corrective actions in real time.
Example Scenario
A shipment is delayed at a port and requires intervention.
Expected Capabilities
- Real-time shipment tracking.
- ETA prediction and delay detection.
- Recommendation of corrective actions such as rerouting or vendor switching.
Technical Requirements
- Backend API (FastAPI recommended, but not mandatory).
- Frontend dashboard (React recommended, but not mandatory).
- Continuous state evaluation, action triggering and agentic log trace.
Problem Statement
Build an AI agent capable of executing real-world workflows from natural language commands.
Example Inputs
- "Analyze this CSV and provide top trends"
- "Schedule a meeting tomorrow"
- "Summarize and send today's reports"
Expected Capabilities
- Intent recognition.
- Dynamic tool selection.
- Multi-step execution.
- Action logging and traceability.
Technical Requirements
- Backend API (FastAPI recommended, but not mandatory).
- Frontend dashboard (React recommended, but not mandatory).
- Integration of at least three tools with execution tracing.
People
Judging Panel and Mentors
Overseas Jury
Pradeep Natarajan
Associate Vice President of Engineering, AI Innovation - NeoGov (USA)
16+ years in ML and data science. Built high-scale fraud detection and recommendation systems. Focused on custom LLMs and multi-agent validation frameworks.
Logeswaran Audhikesavan
Senior Director of Engineering - Deepfrog AI Solutions Pvt. Ltd. (UAE)
21+ years in AI architecture, enterprise platforms, and distributed systems. Leads production-grade agentic AI ecosystems in logistics and finance domains.
Industry Mentors
Iyappan S
Staff Machine Learning Engineer - Deepfrog AI Solutions Pvt. Ltd.
Builds scalable AI ecosystems with NLP, forecasting, and deployment pipelines on AWS and Azure using robust feedback and analytics loops.
Viknesh S
AI and ML Engineer - Deepfrog AI Solutions Pvt. Ltd.
Specializes in RAG, autonomous agent architectures, and LLM optimization across logistics, healthcare, and remote sensing applications.
Venue and Operations
Logistics designed for 24-hour execution
Infrastructure and Technical Setup
- High-speed internet and continuous power backup.
- Projector, AV support, and remote judging setup.
- Extension boards and team-ready seating.
Participant Support
- Dedicated hacking area, mentoring corners, and presentation zone.
- Refreshments throughout the event duration.
- OD support and basic first aid availability.
Scoring
Criteria
Agent architecture and system design quality
Tool integration and orchestration strategy
Reliability, edge-case handling, and testing depth
Output quality, explainability, and observability
Code quality and implementation discipline
User experience and clarity of demonstration
Event Information
Everything you need to participate
Dates
April 16 (10:00 AM) to April 17 (10:00 AM)
Mode
Offline / In-Person
Team Size
2-3 members
Eligibility
Open to 2nd and 3rd year students
Submission
GitHub repo, working demo, logs/traces, and README
Fee
Rs.100 per team
Schedule and flow are indicative and may be optimized during event operations.
Contact
Reach out for registrations and queries
SRM Point of Contact
Mukesh
+91 755 006 6109Deepfrog AI Point of Contact
viknesh@deepfrog.ai +91 97899 97844Registration
Secure your slot early. Limited team intake for high-quality mentoring.
Register Now