AI Engineer · building reliable AI agents // AI workers
I build agentic systems that go beyond impressive demos: tool-calling, retrieval, streaming UX, voice interfaces, and service boundaries that hold up in production.
My current focus is agent reliability: making AI agents easier to operate, measure, debug, and trust when they hit real users, real tools, and real workflows.
- LangGraph and OpenAI Agents SDK
- RAG pipelines with Postgres, pgvector, hybrid search, and evaluation
- Streaming agent interfaces with FastAPI, Next.js, and the Vercel AI SDK
- Voice AI agents with LiveKit and real-time orchestration
Taking those agents from demo to dependable: evals, guardrails, durable execution, observability, and human-in-the-loop on the actions that touch real data or money.
- Contributed to LangGraph Open Canvas
- Built AI agent systems for multiple startups in 2025
- Specialty: Multi-agent workflows, canvas-style UX, voice AI integration
- Canvas Callback - Canvas-style AI interface (80+ stars)
- LangGraph Voice Call Agent - Real-time call to a LangGraph agent over LiveKit
- OpenAI Agents Streaming API - Stream responses from openai-agents-python using fastapi
- postbot3000 - (280+ stars)
AI agents are becoming digital workers, but the useful version is not magic autonomy. It is scoped tools, measured quality, reliable execution, approval gates where they matter, and systems a team can replay and debug after something goes wrong. That is the layer I care about building.
Paperclip · LangGraph · OpenAI Agents SDK · Vercel AI SDK · RAG · pgvector · Python · FastAPI · SQLModel · PostgreSQL · Next.js · TypeScript · React · Docker
Open to selected freelance and consulting work around reliable AI agents, AI workers, and agentic product systems.
📧 ahmadshaukat_4@outlook.com
💼 LinkedIn
🐦 X/Twitter
