[email protected] | (650) 549-5075

the Build Maestro

the Build Maestro

I build production AI systems that run on your own hardware. My current work centers on ambient intelligence: software that listens, classifies intent at the moment data arrives, and acts on it through confidence-gated execution. Underneath that work is two decades of backend, infrastructure, and release engineering across Python, AWS, and the systems that hold them together. Go is what I reach for now.

AI Engineer • Go • Local LLMs • Event-Driven Pipelines • Self-Hosted Infrastructure

Core Focus

AI Systems • Ambient Intelligence • Go • Local LLMs • RAG • Event-Driven Pipelines • Self-Hosted Infrastructure

Building production AI systems and self-hosted infrastructure that run on your own hardware, on top of two decades of backend and platform engineering.

AI Engineering

Production AI systems in Go

Building intelligent systems that classify, reason, and act on data as it arrives.

  • Local LLMs and Whisper-based transcription
  • Semantic action vectors and intent classification
  • Custom RAG pipelines with intent-aware ingestion
  • Multi-agent orchestration and context routing
  • Confidence-gated execution

Self-Hosted Infrastructure

Single-binary Go services for the homelab

Building software you can run on your own hardware, with no runtime dependencies.

  • Single-binary Go deployments
  • Event-driven pipelines with pub/sub
  • SQLite and embedded vector stores
  • Smash Deck dashboard family
  • Tailscale-aware networking

Backend & Release Engineering

Two decades of platform experience

The foundation underneath the AI work: reliable systems that hold up under operational pressure.

  • PostgreSQL backup, restore, and FDW workflows
  • AWS API integrations and cost optimization
  • CI/CD pipelines and release orchestration
  • Systems integration across SaaS and internal tools
  • InstallShield and enterprise distribution

SERVICES & EXPERTISE

What I build and how I solve problems

Ambient Intelligence

Software That Listens, Classifies, and Acts

My current flagship project, Nogura, is an ambient intelligence platform built in Go. It uses semantic action vectors, temporal claim tiering, and confidence-gated execution to act on speech and text as it arrives, instead of waiting for the user to ask the right question. Local LLMs, local vector store, your data stays on your hardware.

Learn More →

Applied AI Systems

Local LLMs, RAG, and Multi-Agent Orchestration

I build applied AI systems that run on your own infrastructure. Custom RAG pipelines, OpenRouter and Anthropic API integrations, multi-agent orchestration, and intent classification at ingestion time rather than retrieval time. The emphasis is on usefulness, correctness, and control over novelty.

Self-Hosted Infrastructure

Smash Deck Family of Dashboards

Smash Deck is a family of single-binary Go services I built for the homelab. UniFi controllers, Wyze devices, PiKVM, Plex, and LG webOS each get their own focused dashboard. No containers, no runtime, no cloud dependencies. Configuration via environment variables, persistence in SQLite or flat files.

Full-Stack Development

Front End + Back End Delivery

I design and build complete web applications across the front end and back end, focusing on clarity, performance, and maintainability. I work primarily in Go, with strong Python experience, and build modern UIs using HTMX and Tailwind.

Learn More →

Data Engineering & Reliability

Databases, Backups, and Recovery

I design bespoke PostgreSQL backup and restore solutions tailored to real operational constraints, including FDW-aware workflows and large-scale data movement. Reliability and recoverability are treated as first-class concerns.

DevOps & Release Automation

Delivery Systems & Cloud Integration

I design and implement build, release, and deployment workflows using direct AWS API integrations and third-party service automation. My focus is on repeatable, auditable delivery pipelines that reduce operational friction.

WORK EXPERIENCE

Contracting also full time inc.

Moody's Analytics

Moody's Analytics

Large-scale cloud automation supporting a SaaS platform used by financial institutions worldwide.

Learn More →
GE Digital

GE Digital

AWS Cost Savings

Learn More →
macys.com

macys.com

DevOps member of "lean startup team" within enterprise. Coding CM tools, administer CI

Learn More →
Expedia

Expedia

Coding for Tooling deployments. Created a self service app for developers. Received glowing reviews from management.

GOOD Technology

GOOD Technology

Brought on as a Sr. Packager extending their software installer to support customer database upgrades and coordinated master/slave deployments within constrained network environments.

Seagull Scientific

Seagull Scientific

Release Engineer and Software Packager. Achieved the impossible: Discovered a method of breaking out of the limited UI and attaching a Winforms process to Installshield executable using C# DTF code.

Learn More →
F|NE D3S|GN

F|NE D3S|GN

I started this venture in 2016 and provide a range of services including web design, audio/video production and editing.

ABOUT ME

The early years; what brought me here

Learn More →

WHAT'S ON MY MIND

My Blog

Read the Blog →

Interested in discussing a role or contract opportunity?

Considering a contract or consulting engagement? → Work with me

[email protected]

LinkedIn

Stack Overflow

About this site

This site itself is a small example of how I work. It's bespoke, designed and built by me end to end, and intentionally kept simple and inspectable. I treat it the same way I treat production systems: clarity over cleverness, explicit behavior, and long-term maintainability.

If you're curious about the technical decisions behind it, I wrote about migrating it from Flask to Go here.

→ Read: Migrating this site from Flask to Go

the Build Maestro