William McKeon— Applied AI Engineer
I build agentic systems, LLM applications, and the backend infrastructure behind them.
I’m a Computer Scientist at the Naval Undersea Warfare Center, a Navy research and development laboratory, where since 2026 I have worked with senior leadership on the organization’s AI roadmap. On my own time I built openagent-code, a self-hosted autonomous coding agent; OpenAgent, a multi-service agent platform; and BoeNet, from-scratch conditional-compute research in PyTorch. I also maintain Arcus Code, a hardened distribution of a 266K-line open-source coding agent. That independent body of work is what led to the AI work I own today.
What I do
Agentic Systems
I build autonomous agents around a plan → act → verify loop with typed tool calling, subagents, cross-session memory, and context compaction. openagent-code is ~27K lines of Python with 25 tools, a permission engine with allow/ask/deny rules and a workspace boundary, and a 23-task evaluation harness with objective per-case verifies.
LLM Applications & RAG
I ship LLM-powered APIs in Python and FastAPI with SSE streaming and provider-agnostic routing through LiteLLM, Amazon Bedrock, and vLLM. CareerAgent is a 14-service vertical product — ~37K lines, 72 endpoints, ~800 hermetic tests — with deterministic anti-fabrication grounding and a fail-closed LLM verifier.
Backend & Infrastructure
I design microservices and REST APIs in Python, Go, C++, and TypeScript, packaged with Docker and shipped through CI/CD. OpenAgent is a five-service platform at 135/135 tests passing, running on AWS and self-hosted RunPod GPU infrastructure behind an identity gateway and a model-routing proxy.
- 7 years engineering production software
- Active DoD Secret clearance
- Open source, Apache-2.0
- B.S. Computer Science, URI
Background
Navy research engineering, and the agent work that grew out of it.
I have been a Computer Scientist at the Naval Undersea Warfare Center since 2019, doing software and computer-science work for Navy research and development programs. Much of that work was machine learning applied to sensor and signal data — classification, clustering, dimensionality reduction, and anomaly detection for signal analysis and automated decision support. I guided a team of Navy engineers on an ML data-analytics application that we demonstrated to government stakeholders.
Since 2026 I have worked with senior leadership on the organization’s AI roadmap: identifying where AI genuinely applies, assessing feasibility, helping set technical direction, and developing the briefings that translate AI capability into concrete applications. I also build the codebase for those applications, so the recommendation and the working software come from the same hands.
I hold a B.S. in Computer Science from the University of Rhode Island, earned in 2022.
How I think about this
What I believe after building these systems.
- 01
The permission engine matters more than the orchestration loop.
Deciding when a human belongs in the loop is what separates an agent people trust with real work from one they quietly stop using.
- 02
The dataset and the evaluation decide whether it works.
Far more often than the model does. I write the eval and the verifies before I tune anything, so improvement is something I can measure rather than something I can feel.
- 03
Build from first principles, then reach for a framework.
I write the transformer, the router, the loop by hand at least once, because you cannot debug what you do not understand.
- 04
Own it end to end.
Architecture, data model, deployment, monitoring. The parts nobody owns are the parts that fail in production.
- 05
Document honestly, including the limitations.
I write for someone who did not build the system, and I say plainly where it stops working.
Common questions
The things people ask first.
- What kind of role are you looking for?
- I work as an Applied AI Engineer on agentic systems, LLM applications, and the backend infrastructure behind them. What I want is ownership of systems end to end, from architecture through evaluation.
- What have you built?
- Six systems, described in detail on the Projects page: openagent-code, a self-hosted autonomous coding agent; OpenAgent, a five-service AI platform; CareerAgent, a 14-service vertical product; BoeNet, from-scratch PyTorch conditional-compute research; Arcus Code, a hardened distribution of a 266,000-line open-source coding agent; and DatasetForge, a TB-scale LM-pretraining data pipeline.
- What is your day job?
- I have been a Computer Scientist at the Naval Undersea Warfare Center since 2019, doing software and computer-science work for Navy research and development programs. Since 2026 I have worked with senior leadership on the organization’s AI roadmap and built the codebase for the AI applications behind it.
- What machine-learning work have you done outside of agents?
- I have applied supervised and unsupervised methods to sensor and signal data — classification, clustering, dimensionality reduction, and anomaly detection — for signal analysis and automated decision support, and I guided a team of Navy engineers on an ML data-analytics application demonstrated to government stakeholders.
- What are you working on now?
- Independently I continue to develop openagent-code and its LoRA-SFT distillation pipeline, including the automated base-vs-student promotion gate, and I maintain Arcus Code. At work I continue to build the AI applications behind the organization’s roadmap.