Frontier Engineering Challenge 2026

About

AI can produce convincing code in seconds. Real engineering begins when convincing is not enough: incomplete requirements, hidden dependencies, difficult edge cases, failure modes, and decisions that require technical judgment. The micro1 Frontier Engineering Challenge is a free, global, online competition for engineers ready to work at that frontier. Over a three-day sprint you will use coding agents to tackle a real-world software engineering problem and produce a solution that is correct, reproducible, testable and clearly explained. The full challenge statement is released at kickoff (Aug 28 at 15:00 UTC). Until then, expect to build at the frontier of agentic AI without details that would reveal the problem in advance. Participants will be able to registre after the Hackathon kicks off. The Hackathon ends on Aug 31 at 18:00 UTC. Participants are evaluated on the quality of the submitted work, not résumé prestige, job title, university, employer or geography. Coding-agent use is required. You must disclose the tools you used and submit the required trajectories for evaluation. Theme Build at the frontier of agentic AI. Use coding agents to tackle a real-world engineering problem where correctness, reproducibility and human judgment matter. In short: let your creativity run. Show us how you think through code. Apply it to any industry you like. The full problem PDF, starter materials, constraints and acceptance tests are published at kickoff. Every valid entry must present both a baseline solution and an advanced solution. The advanced solution should show a meaningful improvement in capability, reliability, efficiency, coverage or engineering quality, not a cosmetic variation. Technology policy Supported languages include Python, TypeScript, Java, C++, Go and Rust. Commonly used frameworks and libraries in those ecosystems are allowed, provided the entry stays reproducible and complies with the final problem PDF. Illustrative examples: FastAPI, Flask, Django, LangGraph and related Python ecosystems; Node.js, Express, NestJS and Next.js for TypeScript; Spring Boot for Java; standard C++ and CMake toolchains; Go modules and common Go web frameworks; and Cargo, Tokio, Axum or Actix for Rust. This list is intentionally non-exhaustive. The problem PDF may prescribe a starter repository, runtime, dependency limits, API access or testing environment where needed for fair and deterministic judging.