AI Adoption Specialist
I build production AI systems, then make non-technical teams trust and adopt them. Marketing and commercial background, hands-on engineering, one job: take AI from capability to production and make it stick.
Flagship build
Sole architect & engineer · Cloudflare Workers, TypeScript, Next.js, Supabase, Claude Vision, Voyage, Slack API
Claude Vision writes structured metadata back to the DAM. A two-source merge splits product identity from scene context, validated against 35K ground-truth assets with an automated precision/recall gate.
A leadership-level intelligence layer over the library, product docs, and business knowledge. Ask anything, get a consistent, sourced answer. One source of truth.
Natural-language asset search inside Slack. Type what you need, get images and video from the 500K library. No DAM login, no training.
The same layer answers product questions via Slack from official docs. Consistent answers across sales, marketing, and ops.
"The Slack bot is the adoption story. I did not train the team to use the DAM. I brought the DAM to where they already were."
How I work
Understand the real workflow, ship production-ready fast, then coach until the team's confidence is self-sustaining.
The documented process and the real one are rarely the same. I embed with teams first and design second. That is where adoption risk lives.
I wire AI into the platforms teams already own rather than asking them to adopt something new. Lower friction means higher uptake.
I work alongside people in their real tasks until they are confident. Training teaches tools. Working together builds judgment.
Client & business systems
Odoo · social APIs · Claude API
Clients managed social content manually across separate tools. I built AI pipelines connected directly into Odoo, the platform they already ran on, automating generation, approval, and publishing.
Adoption was immediate because the workflow lived where the team already spent the day.
RSS · web crawlers · Scrape Creators · Claude API
A multi-source system, RSS, web crawlers, and X monitoring, that pulls industry signals and filters them against a business-relevance gate.
A daily digest feeds an implementation queue I use to advise clients on where AI moves the needle.
Stack
The other side
AI adoption fails at the people layer far more often than the technology layer. A human-aligned future of work is built by people who trust the tools, and that trust is designed, not assumed.
I have worked across marketing, sales, and communications teams before moving into AI engineering. That commercial background taught me how to read a room, where resistance comes from, and how to build enough trust that people will try something new in their actual workflow. I understand the business side, not just the build.
What I have learned: the fastest path to adoption is making someone feel successful with the tool in the first five minutes. Everything I design is built backwards from that moment.
Trusted by
Let's build