Two kinds of work. Products we conceived and built ourselves, and solutions delivered inside client organizations — shared at a level those clients are comfortable with.
Our own products — conceived, designed, and architected in-house, from problem statement to shipping system.
From a hackathon problem statement I drafted to a shipping AI team brain — product design, roadmap, and UI/UX, then the architecture of an agentic layer that plans, uses tools, and acts inside a permission-scoped space model.
Natural-language analytics on enterprise data, built on Controlled Agency — a bounded, auditable pipeline for the common path, with true open-ended agency isolated to a hard-capped premium tier.
Built inside client organizations, across enterprise cloud, applied GenAI, and data engineering. Anonymized where needed.
Acting Head of Engineering for a new-age veterinary care company — building the team, platform, and digital experience that scaled the business from one clinic to 80+.
A companion to the veterinary platform — a single-purpose, guardrailed booking agent that added net-new bookings without cannibalizing the existing workflow.
A companion to the veterinary platform — a medallion-architecture data warehouse that unified payments, clinical, and practice-management data from three source systems into one governed source of truth for BI.
Lead Architect on a full-stack RAG solution — built in the GPT-3.5 era — that cut training-content creation from 5 days to under 15 minutes and gave sales agents a live AI co-pilot on customer calls.