MAPPy
An LLM + RAG platform that helps NASA and DoD teams generate and reason about systems engineering artifacts
I lead backend development for MAPPy at Booz Allen Hamilton. It’s a platform that helps systems engineers at NASA and across the DoD generate, review, and reason about their artifacts — requirements, trade studies, architecture descriptions — with large language models, instead of doing all of it by hand.
Most of my work is on the server that does the actual AI processing: the retrieval-augmented generation (RAG) pipelines that pull the right context into a prompt, the orchestration around the model, and the plumbing that keeps the whole thing reliable. I also own our deployments, which are Dockerized and run in secure environments like NASA’s AWS GovCloud. A good chunk of the job is getting generative AI to behave inside the constraints that come with a government cloud.
I’ve spent a fair amount of time on the other side of the screen too — demoing MAPPy and walking through use cases with engineers and stakeholders at NASA and DoD facilities.
MAPPy is a government project, so it isn’t open source.