I design and ship production AI systems that transform complex documents and data into reliable decisions.

AI Engineer focused on LLM systems, autonomous agents, and data-intensive AI platforms. I work across the full lifecycle: architecture, retrieval engineering, orchestration, observability, and production deployment.

My specialization is building high-impact AI solutions in regulated and document-heavy environments, combining:

• RAG architectures (retrieval quality, chunking strategy, embeddings, vector search)
• Agentic workflows (tool use, memory, stateful orchestration)
• Data/ML engineering foundations (pipelines, reliability, performance, governance)

Jul 2025 – Present

• Leading the implementation of AI initiatives across public-sector business areas, including the institution’s first large-scale AI programs.
• Co-developed an intelligent legal/administrative assistant with LangChain, LangGraph, and LlamaIndex, integrating RAG, OCR, and memory-enabled agents.
• Built AI auditing agents that analyze approximately 1M documents per year, enabling complete annual processing for the first time in the institution’s history.

Feb 2025 – Present

• Delivering advanced workshops and bootcamps covering RAG design, Transformer foundations, and LLM implementation for business use cases.
• Producing technical content and mentoring engineers through practical projects in semantic search, vector databases, and intelligent data workflows.

• ai-engineer-roadmap Structured learning path for AI Engineers, from fundamentals to production architecture.

• workshop-ai-agent Practical workshop repository focused on building AI agents with orchestration, tools, and real-world workflows.

• rag-project End-to-end RAG pipeline using LangChain, LangGraph, Qdrant, and Langfuse with a production-oriented architecture.

• Retrieval-Augmented Generation (RAG) architecture and optimization
• Agent design with tool calling, memory, and multi-step reasoning flows
• Prompt/system design for robustness, consistency, and safety

• Chunking strategy design (semantic, fixed, and hybrid)
• Embedding model evaluation and retrieval quality tuning
• Vector search pipelines with hybrid retrieval and reranking patterns

• ETL/ELT for AI-ready datasets (batch and near real-time)
• API-first backend services for AI products
• Monitoring and evaluation for LLM applications (latency, quality, traceability)

• Deployment patterns for cloud-native AI systems
• Cost/performance tradeoff analysis for model and infrastructure choices
• Reliability and observability in mission-critical AI workflows

LangChain · LangGraph · LlamaIndex · OpenAI API · Qdrant · Langfuse

Python · FastAPI · Pydantic · Uvicorn

Pandas · NumPy · dbt · Apache Airflow

PostgreSQL · MongoDB · Redis · Snowflake

AWS · GCP · Azure · Docker · Git · Linux

• Advancing agentic AI systems for large-scale document intelligence in the public sector
• Mentoring engineers to move from data foundations into production AI engineering
• Building reusable patterns for trustworthy, observable, and scalable LLM applications

Open to collaborations on AI Engineering, RAG platforms, and agent-based automation systems.