Real clients. Real numbers.
A sample of what's shipped. Every one started as someone's manual, painful process.
A 165k-row sheet was choking on array-formula recalculation. Rebuilt as an offline Python pipeline that pushes clean data back through the Sheets API.
Manual payroll calculations replaced with an Apps Script routine that reads hours and outputs pay, every cycle, with no hand-math.
Public sources scraped and AI-classified into a clean, categorised member directory, with no manual data entry.
A scraper watches competitor listings on a schedule and fires an alert the moment a relevant price moves.
A legacy Apps Script project migrated from the deprecated Rhino runtime to V8 and consolidated from 21 files into one maintainable codebase.
A reusable Google Sheets model that runs every real-estate deal through the same assumptions, scenarios and outputs.
A ten-sheet Excel operations workbook with CSE array formulas and a live dashboard, built under a strict no-VBA, no-Power-BI constraint.
A serverless pipeline that takes a phone call, transcribes it, extracts the donation details with an LLM, and writes a structured row to Google Sheets.
A 1,398-record competitor dataset built, QA'd and corrected, then turned into an interactive Leaflet map and a portfolio infographic.
A repeatable pipeline for taking exported web apps to a live, password-gated deployment on Vercel — diagnosing the build failures most people get stuck on.
Shopify stock, supplier feeds and a master Google Sheet kept in continuous two-way sync, so inventory is never wrong twice.
A real-time React dispatch board on Supabase that shows every job and driver status as it changes, replacing a physical whiteboard and a group chat.
Scraping plus enrichment plus sequenced cold outreach, assembled into a lead engine that surfaces qualified prospects every week on autopilot.
A web portal on Supabase + React giving owners and tenants their own secure logins for documents, requests and status — replacing scattered email threads.
A Power Query–driven Excel pack that ingests the source files and rebuilds the month-end reporting automatically — refresh instead of rebuild.
A booking-to-calendar automation that creates events, sends confirmations and fires reminders automatically, cutting no-shows without manual follow-up.
A data pipeline feeding a single Looker Studio dashboard so every store's KPIs sit side by side and refresh on their own.
A Python + OCR pipeline that reads incoming invoices, extracts the fields and writes a structured row straight into the ledger sheet.
A scraping-and-alerting system that monitors competitor pricing across SKUs and flags meaningful moves, with a dashboard for the full picture.
A web app with authentication and a paywalled member area, deployed on Vercel — content behind a clean, reliable login.
Have a process that looks like one of these?
Describe it in one sentence. You leave the call with a clear scope and a fixed price — not a sales pitch.