
Building AI products and third-party integrations for a multi-tenant clinic CRM (Python/FastAPI, TypeScript/Node.js, PostgreSQL).
- AI chatbot, sole author: production RAG chatbot (Python/FastAPI) that answers patients over REST and WebSocket, with hybrid retrieval on PostgreSQL + pgvector, grounded LLM responses, in-chat appointment booking, and human handoff; multi-tenant and bilingual (English/Arabic).
- Voice AI agent, sole author: real-time agent that books, reschedules, and cancels appointments by phone (Twilio) or browser, streaming to a realtime LLM (Google Gemini Live / OpenAI Realtime) with tool-calling that triggers live CRM actions.
- Meta integration hub, primary engineer: WhatsApp Business Cloud API team inbox, Facebook Messenger & Instagram, and Meta Lead Ads that auto-create CRM leads, all on the Graph API.
- Meta Ad Studio: native ad creation on the Marketing API, plus ownership of the Meta App / Business Manager setup: OAuth onboarding (Embedded Signup + Facebook Login), webhook subscriptions, and signature verification.
- CRM, full-stack contributor: work across the SaaS CRM (Node.js/Express + Prisma/PostgreSQL + React), including relational migrations, a UTM marketing-attribution engine, and a visual automation flow builder.
- Cloud & delivery: containerised the AI services with Docker and shipped them to AWS (EKS via Helm, ECR images, S3 call recordings) through GitHub Actions CI/CD.
- Python
- FastAPI
- TypeScript
- Node.js
- PostgreSQL
- AWS

