Ahmad G.

I'm a senior software engineer with around 9 years building production systems, now focused on AI. I build AI agents, RAG systems, and the backend that keeps them reliable in production, not just in a demo. Most "AI developers" can wire up a chatbot. I'm an engineer and solution architect who has shipped systems at real scale. So I handle the hard parts: keeping answers accurate and grounded, controlling token cost and latency, isolating data per tenant, and wiring AI into the tools a business already uses. What I build: AI agents, chat and voice, that answer, qualify, and take action RAG systems: chunking, embeddings, hybrid search, evaluation, and hallucination guardrails Real-time voice agents (streaming, low latency) Custom API and webhook integrations, and two-way syncs between systems Scalable backends: APIs, event-driven services, background workers, caching, observability My stack: Models: OpenAI, Claude, Llama, Amazon Bedrock AI: LangChain, LangGraph, Crew AI, MCP tool servers, vector databases (pgvector, Pinecone, and similar) Backend: Python, Node.js, REST and WebSocket APIs, PostgreSQL, event-driven (Kafka, queues), Docker, AWS and Azure I work with Claude Code and MCP daily I've delivered production systems across banking, telecom, aviation, and large enterprises, as a solution architect and team lead. So I care about clean architecture, predictable cost, and systems that hold under load. If you're building an AI product or backend and want it done right the first time, send me the details. I'll tell you honestly how I'd approach it.