Where I learned "personalization" is just math nobody wants to do by hand.
Grew the paying user base ~27% month-on-month, lifted engagement 44% via RFM segmentation, and wired up CRM automation that quietly improved retention by 42%.
Hi, I'm Taha.
Data analyst by trade. AI-agent wrangler by 2025 necessity. This page is my resume, minus the boring font.
The Journey So Far
Handed a $5M budget and told to "make it work." Cohort analysis beats vibes, every time.
Won back 55,000 average monthly buyers by chasing down a retention drop, and built a daily active-user forecasting model accurate to ~91%.
Studied LLMs and transformers in 2022 — before ChatGPT made it cool to have an opinion about them.
Operational Research & Risk Analysis, graduated with Distinction.
Found £350K hiding in a spreadsheet nobody wanted to double-check. You're welcome.
Built an AI-powered tech news tool with Gemini that cut research effort 77%, and automated team workflows saving ~150 hours a quarter.
Fired Tableau. Hired an AI agent instead. No complaints filed yet.
Replaced Tableau with a vibe-coded analytics platform on Cloudflare Workers, Workers KV, and Analytics Engine — built with OpenCode and an AGENTS.md context layer so AI agents contribute reliably. Helped Strategy hit a 92% ACV target across 9 verticals, and automated reporting to claw back ~140 hours a quarter.
Tools I Reach For
SQL
Python
BigQuery
Power BI
Tableau
Cloudflare Workers
Workers KV
Analytics Engine
OpenCode
Windsurf
Alteryx
Google Apps Script
CI/CD
Looker Studio
Somewhere between chasing paid-media budgets and teaching AI agents to ship production code, I picked up a weirdly specific skill: I can smell a bad dashboard from across the room, and I'd rather prototype something real than sit through another slide deck about "AI strategy." This site is Exhibit A.
Connect with me on LinkedIn →