Projects
AI systems I have built for common business problems. Each one shows the problem, how it works, and the public code.
Sales reps spent hours sorting leads that were never going to buy.
Researches, scores and checks every new lead, then sends one daily list of who to call first.
Invoices and receipts were typed in by hand.
Reads PDFs and phone photos, checks the totals, and returns clean, structured data.
Support teams were buried in routine calls and chats.
Answers voice and chat in one place, updates the CRM, and hands over to a person when needed.
WhatsApp enquiries were answered late, or never.
Replies to common questions straight away and logs every conversation in the CRM.
Nobody had time to keep an eye on competitors.
Checks competitor websites every week, spots what changed, and sends a short brief.
How every project page is written
The problem, in the words a manager would use.
The architecture, in a diagram you can trace.
The result, with code you can open in GitHub.
What we learned: the hard part, not the victory lap.