Three platforms live in production, written solo: a multi-tenant AI customer-operations platform, a community for German learners, and a learning system I built to teach myself the discipline. Row-level security, migrations, evals, payment webhooks — the parts nobody films.
This week · Finishing Franco-Arabic language handling in the TalosOps agent, and starting a video series on building real software with coding agents.
Platforms live
Lines TypeScript
Tests
Migrations
Commits in 2026
Read from origin/master on 7 Sept 2026 across three repositories. Not estimated.
Selected work
Real product screens, real numbers, and — on every case study — a section on what is not true yet.
Multi-tenant AI customer-operations platform
One platform running customer service for many businesses at once — an agent that answers and acts across every channel, and Talos, an operator the owner talks to in Egyptian Arabic that runs the platform for them.
A community platform for learning German
A feed pitched at your level, courses, AI pronunciation practice, full telc/Goethe/ÖSD mock exams with a fresh sheet every time, publish-first moderation, and Egyptian card payments.
An evidence-based learning platform, built for one student
A full applied-AI curriculum — 16 phases, 120 lessons, written not generated — with spaced review, a tutor that gives hints before answers, and hand-written Arabic on every single item.
Inside the systems
Interactive on the case-study pages: ask Talos, search knowledge across languages, break a ranking function.
94 actions, two speeds — approve every change, or let it run and record — and one-approval plans. Runs the platform for the owner, in Egyptian Arabic.
openAnswers in the customer's dialect from the business's own knowledge and live data, changes orders only with confirmation and verified identity, and hands off to humans.
openA question in English or Franco-Arabic finds the right Arabic document — across Notion, Drive, Sheets, Confluence, Miro, files and page images.
openWhatsApp, Instagram, Messenger, Telegram, Gmail, web chat; Shopify, Paymob, four carriers, four CRMs, calendars, an OpenAPI connector and a local bridge.
openEvery closed conversation evaluated, root causes clustered, a simulation harness before real customers, and a read-only quality copilot.
openBounded signals, decaying interests, publish-first moderation, and exams with a fresh sheet every time.
openWriting
Generating code stopped being the hard part a while ago. Everything I do differently now is about the other half: making it cheap to find out whether the code is wrong.
A feed score with a small weight on engagement can still be almost entirely engagement. Bounding every term is what makes the numbers in the formula mean anything at all.
Every moderation design starts with the same question, and almost everyone answers it backwards: does review let content up, or take it down?
Video series
Long-form on YouTube, cuts on Instagram and TikTok, vlogs from inside the work. Including the sessions where the agent confidently ships something broken and I have to catch it — those are the useful ones. First episode in production.