Neo

Neo is a fully autonomous AI engineering agent designed to help ML engineers and researchers build, test, and improve AI systems from natural language task prompts. It automates the machine learning development workflow by researching approaches, writing code, running experiments, debugging failures, benchmarking models, and generating reports. The tool integrates directly into VS Code and Cursor IDEs, allowing users to manage ML tasks through an interactive chat interface while Neo handles iterative engineering work autonomously. Neo excels at tasks including model fine-tuning, prompt optimization, RAG pipeline development, LLM evaluation, and dataset preparation. It can run experiments for extended periods, automatically selecting optimal models and generating versioned artifacts for review. The tool is built for applied AI builders, product managers, and ML engineers who want to accelerate development cycles and reduce manual engineering overhead. A notable limitation is that its capabilities depend on GPU availability, as Neo runs on user-provided GPU cloud infrastructure rather than being fully cloud-hosted.