Backend engineer who ships production AI/LLM systems. I build backend services with Kotlin/Ktor and Python, and I bring LLMs (RAG, agentic workflows, local model deployment) into real production environments — not just prototypes.
- Backend systems — Domain-Driven Design & Clean Architecture, Kotlin/Ktor and Python/FastAPI, PostgreSQL, Kafka, event-driven microservices
- Applied AI engineering — RAG pipelines, agentic workflows (LangGraph/LangChain), fine-tuning, and local/quantized model deployment (llama.cpp, GGUF) when cost or data constraints rule out hosted APIs
Most of my production work lives in private company repos (client and employer codebases), so my public write-ups below are the clearest way to see how I actually work.
- Enforcing types in the Domain layer (DDD) — a practical, pure-Python approach
- Deploying a GGUF Model with CPU using Llama.cpp and Open WebUI
- Fine-tuning, Monitoring and Serving a LLaMA 8B Instruct LLM with a Custom Dataset
- Supercharging OpenAI: Unlocking Speed with Parallel API Calls
A multi-agent system using LangGraph with MCP-based tool servers — in progress, repo coming soon.
- LinkedIn: linkedin.com/in/eghdami