Francisco Ingham is a founder and CTO with nine years of experience building ML-first products and tooling, currently making farm work LLM-native at FieldData. He combines startup grit with enterprise impact from multiple roles at Mercado Libre, where he led squads that shipped production embeddings, AutoML tools, and ML ranking systems that materially boosted business metrics. A hands-on back-end engineer, he contributes to prominent open-source projects like LangChain and Instructor, improving LLM chains, prompts, and integrations for real-world use. He founded Pampa Labs to consult on conversational agents and high-performance semantic search, blending research-driven experimentation with production engineering. Based in Vicente López, Argentina, he brings a rare mix of product leadership, deep NLP/embeddings expertise, and reproducible open-source contributions.
9 years of coding experience
5 years of employment as a software developer
Undergraduate Business Economics, Undergraduate Business Economics at Universidad Torcuato Di Tella
Contributions:2 reviews, 7 PRs, 11 comments in 7 months
Contributions summary:Francisco primarily focused on improving the `instructor-ai/instructor` repository by fixing and improving core functionality. They fixed completion endpoints and made adjustments to readme details within the DSL example, specifically modifying code related to `openai_function_call/dsl/completion.py`. Additionally, the user made adjustments to the tutorials and added a new validation tutorial demonstrating their understanding of the project's concepts. The user demonstrates a good grasp of the core principles of the project.
Contributions:3 reviews, 11 commits, 29 PRs in 1 month
Contributions summary:Francisco primarily contributed to the Langchain library by implementing features, fixing bugs, and improving prompts. They modified the code related to LLM chains, API interactions, SQL database integration, and agents. Key contributions include enhancing prompts for various use cases and implementing fixes, such as addressing issues in Qdrant vector store and SQL prompt compatibility. They also added features to customize verbose and memory configurations, demonstrating their focus on improving the library's usability and functionality.
composability
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.