Senior AI Engineer & Aspiring AI System Architect based in Paris. Passionate about turning cutting-edge agentic research into resilient, scalable, and production-grade distributed systems.
- 🔭 What I Do: Build end-to-end GenAI platforms, high-throughput backend services, and multi-agent workflows with a strong focus on system design and architectural trade-offs.
- ⚙️ Current Focus:
- Scalable Agentic Systems with Model Context Protocol (MCP) and LangGraph / CrewAI.
- Advanced Retrieval-Augmented Generation (Hybrid RAG, SELF-RAG, Vector DBs).
- High-performance asynchronous backend services with FastAPI, uv, and cloud-native infrastructure on AWS.
- 🌱 Continuous Learning: Deepening knowledge in Distributed System Design, LLM Evaluation Harnesses (LLM-as-a-judge), and cost/latency optimization for inference pipelines.
- 🤝 Open for Collaboration: I love discussing system trade-offs, architecture decisions, and open-source GenAI tooling. Feel free to ping me for tech chats or feedback!
| Domain | Technologies & Tools |
|---|---|
| GenAI & Agents | Model Context Protocol (MCP), LangGraph, CrewAI, Qdrant, RAG Pipelines, LoRA / PEFT |
| Backend & Systems | Python 3.12+, FastAPI, Pydantic v2, MyPy, uv, REST APIs, SQLAlchemy, Redis |
| Cloud & Architecture | AWS (SageMaker, Bedrock, S3, Lambda), Microservices vs Monolith trade-offs, Event-driven |
| DevOps & Standards | Docker, Git/GitHub Actions, CI/CD, Prometheus, Grafana, Clean Code |
- mcp-arxiv-demo: Rich-context AI workflow using Anthropic’s Model Context Protocol (MCP) to query and synthesize scientific papers (deployed live on Render).
- spark-and-python: Distributed data processing workflows, ETL pipelines, and large-scale data manipulation using Apache Spark (PySpark).
- Enterprise Multi-Agent RAG: Scalable architecture integrating vector search (Qdrant), keyword search, and coordinated agent communication via MCP for enterprise document reasoning.
- 💼 LinkedIn: /in/tien-le
- 📧 Email: letien.fr@gmail.com
- 😄 Pronouns: He/Him
- ⚡ Fun Fact: When I'm not profiling LLM latency or refining system diagrams, I enjoy exploring architectural case studies and good coffee in Paris!