Lead Engineer/Programmer with Python Expertise in Habibot

Looking for a Founding Engineer / Programmer (equity) — Habibot Habibot is a chat-first booking assistant (Telegram first; WhatsApp / voice next): find a place → book → confirm → remind — no app download, multilingual. Not a greenfield build. A large codebase already exists. I need someone who can own shipping the next product milestones — finish what’s real, wire what’s missing, and keep security serious. What you get • Equity / profit share on commercial traction • Ownership of architecture, reliability, and what ships • If nothing commercializes: small expense reimbursement (receipts-based stipend; capped — agreed before starting) What I’m looking for • Strong Python (production services + async) • Comfortable with bots, APIs, webhooks, shipping fast • Can extend an existing system — not rewrite for sport • Bias to ownership over ticket closing • Part-time OK at the start; intensity ramps with traction • Nice-to-have: aiogram/Telegram, P2P (Hyperswarm), multi-channel outreach, Telnyx/Twilio CURRENT SOFTWARE STATE (honest) Stack: Python bot (bot/, aiogram) is the active product; older Bun/TypeScript MVP still in repo for WhatsApp/calendar pieces — don’t rewrite unless we decide together. Already strong: booking state machine + slot-lock · group dining (“tavolata”) scaffolding · menu-in-chat pipeline · notifications router · signed ledger · multilingual layer · large test / acceptance suite. Still incomplete: vendor delivery loop (requests can stay REQUESTING) · discovery can fall back to placeholders without live sources · voice/STT and Telnyx not fully live · Hyperswarm / P2P booking mirror still a stub · price/marketing channels mostly not connected. Bottom line: core logic is largely built. The work is closing live loops, growth surfaces, security replication, then multi-vertical booking. MILESTONES M1 — Working restaurant MVP (pilot city) End-to-end on Telegram: real discovery → request table → venue confirm/decline/counter → confirmation + reminder. No stuck REQUESTING. Soft pilot with real venues and real diners. Done when: a diner can book a restaurant and get a real confirmation; venue can answer; reminders fire once. M2 — Tavolata at city density + conversation marketing + Hyperswarm security When a city has enough users: tavolata (group quorum → book together) works reliably in production. Growth layer: enter real conversations where people already talk about going out (e.g. someone comments under a restaurant post → invite them to book via Habibot), across platforms we can operate on. Security layer: use Hyperswarm so bookings are stored/replicated in a P2P-backed path (integrity + durability beyond a single server) — replace the current stub with a real mirror of reservation records. Done when: group bookings succeed in a live city cohort; outreach → chat → book is measurable; bookings are written to the Hyperswarm-backed store and recoverable/auditable. M3 — Every actually bookable service Same engine, new verticals: barber, football/calcetto pitch, spa, and anything else that is genuinely bookable with a confirmation (no prepaid ticket / medical / commission-work). Parametric discovery + booking flow per vertical — not a new product per category. Done when: at least several non-restaurant verticals complete a real book → confirm path with the same reliability bar as M1. How we’ll work • Async-first; English or Italian OK • Start on M1; M2/M3 unlock with users and pilot proof • IP and equity terms documented before deep work To apply, please answer: Which milestone would you take first, and why? Link to relevant work (GitHub / bots / APIs / P2P preferred) Hours/week for the next 90 days Confirm you’re OK with equity + expense reimbursement only if nothing commercializes Serious builders only. Fixed-price “build my app” gigs — skip.