Available for new projects

Production-grade
GenAI. Built fast.

RAG systems · Agentic AI · LLM Engineering · Fine-tuning

I partner with companies to design, build, and ship Generative AI solutions that work in production — not just demos. From RAG pipelines to multi-agent systems, I bring 20+ years of engineering experience and a PhD to your hardest AI problems.

Samsung Capgemini Neogov SiDi Norus

Services

RAG Systems

End-to-end Retrieval-Augmented Generation pipelines — chunking strategy, embedding selection, vector stores, reranking, and evaluation. Built for accuracy and scale.

  • LangChain / LlamaIndex
  • Qdrant · FAISS · Chroma
  • OpenAI · Claude · Bedrock
🤖

Agentic AI

Multi-agent systems that reason, plan, and act. From single-agent automation to complex orchestration with tool use, memory, and human-in-the-loop workflows.

  • LangGraph · CrewAI · AutoGen
  • MCP · Tool use
  • FastAPI · Python
🧠

LLM Fine-tuning

Domain-specific model adaptation using LoRA and PEFT techniques. Smaller, faster, cheaper models that outperform general-purpose LLMs on your specific tasks.

  • LoRA / PEFT / QLoRA
  • Hugging Face Transformers
  • Azure AI · SageMaker
🗺️

Technical Advisory

Strategic guidance for CTOs and product teams on AI architecture, model selection, build-vs-buy decisions, and GenAI roadmaps.

  • Architecture review
  • Team enablement
  • GenAI roadmapping

Past Work

Ubiminds → Neogov

AI Engineer

2025 – Present

Building agentic AI workflows and production RAG pipelines for enterprise HR software. Integrating GenAI capabilities across existing product surfaces.

LangGraph RAG Agentic AI

Capgemini

AI Engineer

2025

Delivered GenAI solutions for enterprise clients using Azure AI and LangChain. Multi-agent systems for automated document processing and knowledge extraction.

Azure AI Multi-agent LangChain

SiDi / Samsung

AI Engineer + Engineering Manager

2022 – 2025

Led a 30-person ML team shipping AI features to millions of Samsung device users. On-device LLM fine-tuning (LoRA/PEFT), RAG pipelines, and multimodal AI for Galaxy devices.

LoRA / PEFT On-device AI Team Leadership

Norus

Co-founder & CTO

2015 – 2022

Co-founded and led technology for a B2B SaaS platform serving healthcare, energy, and public sector. Grew engineering team from zero to enterprise-grade over 7 years.

SaaS Kubernetes Full-stack

Why work with me

20+
Years of engineering experience
PhD
Industrial Engineering — research-backed approach
30
Engineers led at SiDi / Samsung
POC→
Production — not just prototypes

Fabrício Sperandio, PhD

I'm an AI Engineer and GenAI consultant based in Criciúma, Brazil, working remotely with companies worldwide. With a PhD in Industrial Engineering and 20+ years in software, I've led large ML teams, co-founded a SaaS startup, and shipped AI features used by millions.

I focus exclusively on Generative AI — RAG, agents, fine-tuning — and I work closely with clients to make sure what I build actually ships and scales.

📍 Criciúma, Brazil 🌐 Open to remote 🎓 PhD, Industrial Engineering

Let's scope
your AI project.

Available for GenAI engineering contracts, consulting engagements, and technical advisory. Industries: healthcare, energy, finance, consumer electronics, and public sector.