← back
The Rise of the AI Software Engineer: Jesse Han
Takeaway
Morph's bet: combine static analysis + embeddings + graphs into a queryable code index that both retrieves context for any coding agent and generates fine-tuning data.
Summary
- Morph Labs CEO Jesse Han pitches the 'personal AI software engineer' that augments developers across the SDLC and stays current with their codebase.
- Announces the Morph Code Index — a neurosymbolic code database combining static analysis, vector embeddings, and graph algorithms with a query language for precise retrieval.
- Open-source, with Python bindings to attach to any coding assistant for codebase-wide retrieval context; can also traverse a codebase to generate training data for fine-tuned coding models (they trained Rift Coder 7B this way).
- Rift Coder 7B is positioned as the only open-source code-editing model that runs locally on a laptop/IDE.
coding-agentscode-searchembeddings
Original description
Programming was once called "automatic programming". One day, we'll think about how engineering was once called "AI engineering". Along the way, we will command legions of AI software engineers that adapt, evolve, and know our codebases better than we do. I will unveil what we have been working on at Morph Labs to take us into that future. Recorded & streamed live for the AI Engineer Summit 2023. See the full schedule of talks at https://ai.engineer/summit/schedule & join us at the AI Engineer World's Fair in 2024! Get your tickets today at https://ai.engineer/worlds-fair About Jesse Jesse Han is the CEO and founder of Morph Labs, which is building open-source infrastructure for the personal AI software engineer. He holds a PhD in mathematics and previously worked at OpenAI on theorem proving, synthetic data, and GPT-4.