Senior Software Engineer · Production AI
AI engineering that survives contact with production.
I build the systems around capable models: tools, orchestration, context, evaluation, observability, and the distributed infrastructure that makes an agent dependable. My work spans customer-facing conversational AI, high-volume MCP platforms, and open-source systems for long-context reasoning and durable organizational knowledge.
Capabilities
The model is one component. The product is the whole system.
LLM agents and orchestration
Tool-calling agents, structured outputs, multi-step workflows, model experiments, and operating contracts that keep autonomy inside explicit boundaries.
MCP infrastructure
MCP servers, clients, gateways, federation, and production integrations connecting conversational systems to real services at meaningful scale.
Context and knowledge systems
Source-grounded knowledge, recursive long-context analysis, and durable links between product intent, engineering decisions, implementations, and evidence.
Reliability and evaluation
Observability, load testing, failure isolation, cost controls, evaluations, deployment safety, and feedback loops built for production behavior rather than demo behavior.
Public proof
Systems you can inspect, run, and challenge.
Knowledge infrastructure
The Vault
A persistent, source-grounded knowledge system that gives people and agents focused context while preserving the evidence and decisions behind it.
- product decisions connected to engineering implementation
- immutable evidence and regenerable synthesis
- automated ingest, query, and health workflows
Agent developer tools
Billy's AI Skills
Tested Recursive Language Model plugins for Claude Code and OpenAI Codex, with bounded fanout, dry-run planning, local-model support, and a public demonstration corpus.
Applied AI product
Hugo
A self-hosted Slack reading assistant that curates RSS feeds, ranks technical articles with Claude, summarizes links and threads, and delivers a focused daily digest.
Engineering record
Eleven years of distributed systems behind the AI work.
My AI work is grounded in backend and platform engineering across identity, underwriting, analytics, and conversational systems. I have built with Python, FastAPI, FastMCP, C#, .NET, Kafka, Kubernetes, cloud messaging, and production observability.
That background shapes how I approach agents: explicit contracts, measurable boundaries, evidence from production, and graceful failure. I am based in Tri-Cities, Tennessee and work remotely.
Let's compare notes
Building an AI product that has to work outside the demo?
I am interested in remote senior software engineering roles focused on production AI, agent platforms, MCP infrastructure, and distributed systems.