AI Nexus Hackathon 2026:
AI Nexus Hackathon 2026 is a 24-hour flagship AI innovation challenge, aligned with the AI DevSummit.
Participants will work on advanced, future-focused problem statements rooted in emerging AI trends such as generative AI, intelligent agents, multimodal systems, AI ethics, automation, and real-time decision intelligence.
Theme: AI For Intelligent Systems & Responsible Innovation:
Intelligent AI-Driven Health Assistant for Rural Diagnostics
Design an AI system that assists low-resource healthcare workers in diagnosing common diseases using symptoms, images (skin conditions, scans), and patient history.
The system should provide interpretable suggestions, confidence levels, and next-step recommendations while operating offline or on low-bandwidth environments.
Focus: AI vision, federated learning, interpretability, edge deployment
Privacy-Preserving Personal AI Assistants
Develop a personal AI assistant that can operate locally on a user’s device, perform tasks (schedule, notes, summarization), and guarantee privacy by not using third-party cloud APIs or exposing personal data.
Explore techniques like on-device LLMs, differential privacy, or secure enclaves.
Focus: Edge AI, privacy/ethics, lightweight LLMs
Generative AI For Sustainable Agriculture Support
Build an AI solution that uses satellite and sensor data to generate actionable recommendations for small-scale farmers (e.g., irrigation scheduling, pest detection, yield forecasting).
The model should combine generative insights with predictive modeling.
Focus: Multimodal AI, time series forecasting, environmental impact
AI Agent For Developer Productivity In Cloud Environments
Create an autonomous AI agent that monitors cloud application logs, performance metrics, and alerts to propose optimization actions (e.g., scaling, cost saving, auto-fix suggestions) with minimal human input.
The agent should integrate with cloud platforms and display actionable insights.
Focus: Agents, automation, cloud AI
Bias And Fairness Diagnostics Engine For LLM Outputs
Design an AI tool that analyzes outputs from large language models to detect potential biases, harmful content, or misleading associations.
The tool should provide actionable heatmaps or reports for developers to refine models responsibly.
Focus: AI ethics, governance, fairness metrics
Real-Time Multilingual Communication Platform
Build a real-time communication platform that translates speech and text between underrepresented languages with contextual understanding, correctness, and cultural nuance, powered by AI.
It should support both text and voice channels, enabling seamless cross-language interaction.
Focus: LLM/translation, speech recognition, product integration
Intelligent Education Companion For Personalized Learning
Develop a smart educational assistant that adapts learning pathways based on student responses, strengths, and weaknesses.
Use a mix of NLP, reinforcement learning, and content generation to tailor lessons, practice exercises, and progress reports.
Focus: Adaptive learning, educational AI, explainability
Hackathon Structure:
Phase 1 — Online PPT Submission (Screening Round)
Deliverables: (PPT Format: https://drive.google.com/drive/folders/1pv_K-IjDtm0aDHfG2YK9OeyQUlQL1dAE)
Problem statement understanding
Proposed AI-based solution
System architecture & technology stack
Innovation & differentiation
Feasibility & scalability roadmap
Shortlisting Criteria:
Innovation, technical depth, clarity, feasibility, and impact potential
Phase 2 — 24-Hour On-Campus Hackathon
Activities Include:
Intensive development sprint
Mentorship sessions
Mid-review checkpoints
Final demo & jury pitch
ByteHunt competitive rounds
Teams will present their working prototypes before an academic–industry jury panel.
Note:
Accommodation and food will be provided to shortlisted teams during the final round.