I run an independent AI studio and I'm hands-on end to end: agent design, LLM orchestration, integrations, evals/guardrails, deploy, and the dashboard around it. Before going all-in on AI I spent years as a production software engineer — Staff Engineer on the Data Services team at a top-tier mortgage lender (AmeriHome), and a software engineer who built the eCheckpoint product at RICO Technologies (Vue.js, .NET, SQL Server). Computer Science degree, Cal Lutheran University.
What I can build for you: • AI agents that take real work off your team — inbound/outbound phone + chat, lead qualification, booking, follow-ups, document intake, research, and triage • LLM apps — RAG over your own data, grounded chatbots that don't hallucinate, classification/extraction pipelines • Automation & integrations — connect your CRM, inbox, database, and third-party APIs so a manual process runs itself, with a human in the loop where it matters • The full build around it — data plumbing, guardrails, deployment, and a clean admin UI
Proof I build what I pitch: • Agentry - An autonomous outreach + bidding engine that runs my own consulting pipeline end to end — it sources leads, scores each one against an ideal-client profile with an LLM, writes personalized multi-channel outreach, detects and triages replies, and discovers contract work across multiple platforms and drafts proposals for it. Multi-agent, scheduled, running unattended in production — Python, the Claude API, serverless cron workers, Postgres/Supabase, and a Next.js review dashboard.
Stack: Python, the Anthropic/Claude API, RAG, tool-use, n8n orchestration, Postgres/Supabase, Next.js, Vue.js, .NET, SQL Server, Stripe.
How I work: • Outcome first — I scope honestly and won't sell you an agent you don't need • Fast, iterative delivery with regular check-ins, so you always know where things stand • I own it end to end and hand you something that works
Have an AI agent, automation, or app in mind? Send me the details and I'll tell you exactly how I'd build it and what it would take.