// 01 · aboutlast updated 2026-05-24
Renan.
Miqueloti
AI Engineer — LLM systems and applied ML.
5 years in applied AI · 2 in agents and RAG pipelines.
// 02 · stack
Stack
Lists ordered by usage frequency. The leading items are part of my day-to-day; the rest come in as the project demands.
Agents & Orchestration
- LangGraph
- LangChain
- MCP
- Tool-use
- Multi-agent
- Memory
- Guardrails
LLM Stack
- Anthropic Claude
- Prompt Engineering
- Embeddings
- Hugging Face
- Fine-tuning
- LoRA
- OpenAI GPT
RAG & Retrieval
- RAG
- Chunking
- Hybrid Search
- Re-ranking
- Weaviate
- pgvector
- Qdrant
ML & Data
- PyTorch
- scikit-learn
- pandas
- NumPy
- Time Series
- SHAP
- Streamlit
- Computer Vision
- CNN
MLOps & Observability
- LangSmith
- Langfuse
- Evals
- Tracing
- Docker
- CI/CD
- Monitoring
Backend & Infra
- Python
- TypeScript
- FastAPI
- Pydantic
- Postgres
- Redis
- Neo4j
- React
- AWS
- GCP
// 03 · projects
Selected work
Public repos and projects in retrieval, applied ML and agent-based systems. Professional projects remain confidential.
rag-chatbot
RAG pipeline with LangGraph orchestrating retrieve → rerank → generate. Hybrid retrieval BM25 + semantic with RRF, cross-encoder re-ranking, FastAPI streaming and an LLM-as-judge eval harness (relevance, faithfulness, completeness).
agents-AI
Runnable project for AI agents: custom MCP server with 4 tools exposed via stdio, LangGraph HITL and multi-provider (Ollama / Claude / OpenAI).
// 04 · contact
Let's talk.
Open to AI Engineering conversations. Mid/senior, contractor (PJ), remote from Brazil.