Abdul R.

I'm an AI engineer who designs and ships production AI agents, voice agents, and chatbots that hold up under real users and real traffic. Not glued-together templates, but systems built in code, tested, and engineered to scale. I'm Abdul Rehman, a full-stack software engineer with 5+ years building AI and automation systems in Python, TypeScript, and the modern LLM stack. What I Build Custom AI agents with real orchestration (LangGraph and the OpenAI / Claude agent SDKs), tool use, memory, and MCP integrations that connect agents to Slack, Google Drive, CRMs, and internal databases. Voice agents built on LiveKit, Vapi, or Retell, with Deepgram speech-to-text, ElevenLabs or Cartesia voices, and telephony (SIP / Twilio), tuned for sub-second latency and natural turn-taking so callers don't feel like they're fighting a robot. Chatbots and RAG systems done properly: chunking, embeddings, vector databases (Pinecone, pgvector), and reranking, so answers stay grounded in your data instead of hallucinated. LLM integration and fine-tuning across Claude, OpenAI, and open models (HuggingFace), including fine-tuning on your own data when an off-the-shelf model isn't enough. Backends and full-stack work in Python (FastAPI) and React / Next.js, with the databases, APIs, auth, and deployment to back them. And workflow automation with n8n, Make .com, and Zapier when a managed workflow is the right call, and custom code when it isn't. What Separates This From a Quick Prototype I build evals so you can measure accuracy, guardrails so the AI behaves safely, observability so you can see what it's doing in production, and latency and cost optimization so it stays fast and affordable as you scale. Recently Shipped A support chatbot that now resolves around 90% of customer questions on its own, freeing the owner from working weekends. An automated CRM that nurtures leads on autopilot so the sales team closes instead of doing data entry. OCR pipelines for EA Sports using AI vision models. How I Work I don't hand off a black box and disappear. I explain the architecture in plain language, document what I build, and stick around after launch to make sure it keeps performing. 5+ years experience | Full-stack AI engineer who builds systems that hold up in production If you're building an AI agent, a voice assistant, or an LLM-powered product, send me a short note on what you're trying to solve and I'll tell you honestly how I'd approach it.