Devpost
Participate in our public hackathons
Devpost for Teams
Access your company's private hackathons
Grow your developer ecosystem and promote your platform
Drive innovation, collaboration, and retention within your organization
By use case
Blog
Insights into hackathon planning and participation
Customer stories
Inspiration from peers and other industry leaders
Planning guides
Best practices for planning online and in-person hackathons
Webinars & events
Upcoming events and on-demand recordings
Help desk
Common questions and support documentation
.
Multi-agent observability copilot that plans SQL + diagnostics runs, routes metrics/chart agents, and explains your telemetry end‑to‑end
An explainable investment research agent that reveals every reasoning step, tool call, and memory update—turning AI decision-making into a fully transparent, auditable process.
Emergentica: Multi-agent emergency call routing and management system that's cheaper and faster than single-AI systems, because in emergencies, every second counts.
Evidentia lets you fine tune the sources behind LLM outputs so every answer is backed by evidence you can trust! We enhance AI adoption in sensitive areas and even reduce token usage by upto 70%!
a.
testing jailbreaking prompts
Gain control and transparency over the context used by your AI when debugging code.
FlowLens allows you to test and red-team your LangGraph agent systems by exposing APIs to individual nodes (i.e, agents, tools), or subsets thereof - without rerunning your entire workflow each time.
We built a tool that tests and visualises whether AI models can detect edits to their own reasoning. It blends behavioural tests with SAE interpretability to measure true introspective awareness.
Smart Guard blocks 92% of jailbreak prompts, is 3x cheaper than other tools, and balances speed vs accuracy by combining embeddings with LLM guards between user input and AI systems.
Evaluating robustness of black-box LLMs using the power of embeddings.
Know your price and the reason behind it. Insurance made transparent
You can't trust a "black box." Our AI is a "glassbox" that logs its entire "Inner Monologue" every thought, action, and observation. Then you can query and debug it using our Judge.
Automated red-teaming for AI agents
A full framework for interactive multi-agent observability, reliability and efficiency.
LLM-based autonomous control that is explainable, safe, deployable without multi-year learning.
A real-time graph visualization tool ( alternative to Langsmith) for analyzing and exploring claims about financial assets (NVDA, TSLA, GOLD) from news and social media sources.
Too many people have been ghosted by companies. Our solution Synapse uses AI to match a company's job description to technical candidates, providing a interpretable reason for acceptance or rejection.
This project implements an AI agent that helps users find fashion rental items across multiple platforms.
Exposing the critical threat of Harmful Semantic Visual Injection (HSVI) that tricks VLA robots into misinterpreting their environment and objective.
Clear Frame is a News Bias & Coverage Analyser Full‑stack app (Flask API + Vite/React frontend) for multi‑source news retrieval, bias coverage analysis, and observability.
Track C - Dear Grandma
Illusion reveals how companies really use your data. Enter any name or URL, and we surface hidden policies, assess risks, and give clear, AI-generated insights with trusted sources.
25 – 48 of 50
