Portfolio

Selected work

Every project here started with a specific problem: an AI demo that needed to ship to production, a workflow that needed structured evaluation, a system that needed audit logging from day one, or a web platform that needed WCAG done right. The work is the proof.

Based in the SF Bay Area and working with Sonoma County clients. Reach us at contact@sonomasolutions.io.

AI & Agentic Systems

MarketBrew Reference Manual documentation screenshot

Knowledge architect, AI engineer

MarketBrew Reference Manual

A structured knowledge base of evidence-tagged atoms, turned into an exhaustive client reference manual.

An exhaustive, client-ready reference manual for MarketBrew (a deep SEO analytics platform), generated from a structured knowledge base. We extracted MarketBrew’s public reference material into evidence-tagged "knowledge atoms" through a parallel multi-model pipeline, then built the manual from that vetted database under strict schema and QA gates.

  • Knowledge architecture
  • Knowledge-atom extraction
  • Multi-model pipeline
  • Technical documentation
Read the case study →
Emissary AI intake assistant interface screenshot

Full-stack AI Engineer

Emissary

Our flagship AI intake assistant, live in production. Ask it about itself.

Sonoma Solutions' flagship AI intake assistant, live in production at emissary.sonomasolutions.io, where you can ask it about itself. Astro 4, React, and TypeScript on the frontend; Node and TypeScript backend modules for database access, a knowledge base, intake email parsing, lead routing, and downstream email integration. It is the productized version of the AI intake systems we build for clients.

  • Astro 4
  • React
  • TypeScript
  • Node.js
Read the case study →
ADA Document Remediation Engine case study cover

AI Engineer

ADA Document Remediation Engine

Multi-critic AI document remediation engine in Python. Audit-logged QA loops, every model decision inspectable.

A production Python pipeline that ingests structured PDF input plus embedded media, classifies each document to a DocumentProfile, and runs text, table, and structure critics in parallel to evaluate the AI-generated remediation. An audit logger diffs between QA loops so every model decision is inspectable. The architecture ports to other image-and-data-to-generated-content pipelines.

  • Python
  • AI document remediation
  • Multi-critic eval
  • Audit logging
Read the case study →
Video Knowledge System case study cover

AI Engineer

Video Knowledge System

Video knowledge extraction with a real eval harness. Node.js + TypeScript on the Anthropic API.

A Node.js and TypeScript pipeline on the Anthropic API for video knowledge extraction. JSON Schema defines the knowledge atom shape with evidence validation on every persisted record. Real eval harness with eval-retrieval-multi, schema-conformance QA, and baseline snapshot scripts for regression detection.

  • TypeScript
  • Node.js
  • Anthropic API
  • JSON Schema
Read the case study →
MarketBrew knowledge base documentation, the source the Oracle Agent is built on

AI Engineer

MarketBrew Oracle Agent

An eval-first agent on top of the MarketBrew knowledge base, with answers cited back to the atoms.

The agent layer built on top of the MarketBrew knowledge base (see the Reference Manual project). An eval-first agentic scaffold on PostgreSQL: contracts and fixtures are defined before generation, and every recommendation is evidence-cited back to the underlying knowledge atoms. The scaffold pattern transfers across client domains.

  • PostgreSQL
  • Anthropic API
  • Node.js
  • Eval-first design
Read the case study →
SRJC Course Outline Archive data table screenshot

Data Engineer

SRJC Course Outline Archive

Turned hundreds of unstructured catalog pages into a queryable archive.

Course outline data at a community college is public but not queryable: it lives in hundreds of individual web pages with no structured export. This pipeline scrapes and normalizes the full Santa Rosa Junior College catalog, extracting course codes, titles, units, objectives, and outcomes into a consistent archive built for curriculum research.

  • Node.js
  • Scraping
  • Structured extraction
Read the case study →

Web & Accessibility

wcag21aa.com accessibility reference library screenshot

Designer, developer

wcag21aa.com

60+ paired pass/fail WCAG 2.1 AA reference pages. A working compliance library built end to end.

Most businesses do not know their site has ADA exposure until they get a demand letter. wcag21aa.com makes WCAG 2.1 AA compliance tangible: 60+ paired pages showing exactly what a failure looks like and what the corrected pattern looks like, side by side, in language a non-developer can act on. Built end to end as a marketing site and a working client resource, with a custom WordPress theme purpose-built for the reference-library use case. The remediation behind the corrected examples is produced by a multi-critic AI engine (a separate project).

  • WordPress
  • Custom theme
  • WCAG 2.1 AA
  • Content architecture
Read the case study →
Book a 20-minute fit call