Portfolio
Shipping products first, nearest the surface where everyone can see them. Further down, the things I built for myself because nobody else was going to.
A classification API that reads a legal matter and scores every document for relevance, so review teams read the useful few percent instead of all of it.
Ranked review queue
The ASR hub. Speech to text as a service for language workflows, from media transcription through to downstream translation.
Speech to text
The internal hub for machine translation work: evaluations, translations, glossaries, and new engine setup, rebuilt for large jobs and a clearer interface.
Bulk operations
Batch evaluation
1,240 segments
Glossary sync
18 glossaries
Engine setup
DE to EN
Org wide AI spend and usage pulled from every provider into one warehouse, plus the product dashboards that make each AI product readable.
AI spend by provider
A gym and habit tracker I built because nothing on the App Store scored effort the way I wanted. Local first, no account, no server.
Wednesday
Let's Try
Push day
Today
Tell it the car and it tells you what the US used market says it is worth, plus how fast that value falls from here.
Estimate
2019 · 62,000 mi · clean title
RAG platform
The retrieval platform behind internal and customer chat, including GLKnowledge and Ask Nova. I support evaluation 2.0, answer quality, and grounding.
PII redaction API
A redaction service that strips sensitive data before content moves through language and legal pipelines. Built end to end, then handed to the owning team.
Examination support
Matter aware examination support: retrieve case documents, classify witness answers, cite the reason. Hackathon London 2026 finalist as a Digital Reef module.
Visual review prototype
A cluster wheel interface for e-discovery, so reviewers see the shape of a whole collection instead of a ranked list. Demoed on a 2,000 email public archive.