I build robust, scalable solutions that bridge the gap between complex data and accessible user experiences. Specializing in automation, AI integration, and system architecture.
Projects that demonstrate the progression from simple automation to complex, intelligent systems.
Hand it the prompt you already use plus a handful of examples where you know the right answer. It rewrites, scores every version against those same examples, and gives you back the one that measured best — usually inside five minutes.
Maps free-form supplier text onto your standardised terminology without leaving Excel. A Python backend does the reasoning — cache first, then fuzzy match, then live web research — and ranks the candidates by confidence before it commits to one.
Reads Gmail attachments, classifies invoices and receipts with AI, files them into dated Google Drive folders, and logs every one to a Sheets ledger. Runs unattended, on a single Google login.
For 25 years, ranked-by-popularity search has quietly decided which science you are allowed to find. The rest is still in the literature — unindexed, unreachable. This is the R&D on getting it back.
Build logs and live streams: training prompts instead of writing them, tuning LLM pipelines, and the n8n workflows underneath. Short, practical, and the systems are always shown running.
I almost stopped using Google. When I do google, I might as well use Perplexity or AI. What does that mean?
I don't just write code; I engineer systems. My work focuses on creating tools that empower users to handle complex data with ease. From automating mundane administrative tasks to leveraging Large Language Models for semantic understanding, I build software that works.
Python, FastAPI, Node.js, PostgreSQL
React, TypeScript, Tailwind, Office.js
