AI Engineer. 22 years in industrial operations.
I build AI systems that make complex knowledge usable — for industrial teams, personal finance, and blockchain strategy.
I'm Juan José Expósito. PhD in Engineering. Certified Python developer. Former Drilling Engineer at Repsol, ADNOC, and Moeve. Now I apply that operational experience to building AI that actually works in the field.
What I work on
Three focused areas. Each with its own audience, its own problems, and its own solutions.
-
AI for Technical Documentation
Engineering companies with knowledge trapped in PDFs, procedures, and manuals. I build RAG systems and AI assistants that turn static documentation into instant, queryable knowledge.
-
AI for Personal Finance
Automating expense classification, financial reporting, and budget visibility. Practical AI flows for individuals and small teams who want to stop doing repetitive financial work manually.
-
Blockchain & Tokenization
Advisory and technical content on blockchain systems, digital assets, and tokenization strategy. For teams evaluating or building on decentralized infrastructure.
Why it matters to have domain expertise
Most AI engineers build generic solutions. I spent 22 years as a Drilling Engineer before writing a single line of Python. That changes what I build and how I think about problems.
-
I understand your operations
I've worked in the environments where this documentation lives — rigs, plants, control rooms. I don't need to be taught the context. I already know what a critical procedure looks like and why it matters.
-
Engineering rigor, not AI hype
PhD in Engineering. Certified Python developer. I approach AI projects the way I approached engineering problems: with rigorous requirements, clear constraints, and measurable results.
-
Production-ready systems
My stack — Python, FastAPI, LangChain, Weaviate, Pinecone, OpenAI, Anthropic — is chosen for reliability and maintainability. Not for demos.
Selected work
-
AI Research Assistant — Quiet Links
RAG system enabling natural language queries across 200+ academic papers. Delivered in 6 weeks, saving an estimated 1 year of solo development.
Industry: Scientific Research & Knowledge Management
-
Intelligent Expense Classifier
AI agent that reduced expense classification time by 90% — 100+ monthly transactions, 30 categories, zero manual errors.
Industry: Financial Operations / Personal Finance
What clients say
-
Tim Gallati — Founder, Quiet Links Library
"Without exaggeration, I learned the intricacies of developing RAG systems in 6 weeks that would have otherwise taken 6 months to 1 year on my own. Juanjo's deep knowledge of Python development, AI applications, and infrastructure were mind blowing. The proof is in the deliverable — a working AI application in 6 weeks."
Frequently asked questions
Who do you typically work with?
Engineering and industrial companies that need to make technical documentation accessible — oil & gas, energy, manufacturing, utilities. Also individuals and small teams looking to automate financial workflows. And projects related to blockchain infrastructure or digital asset strategy.
What does a typical engagement look like?
Most projects start with a free 30-minute discovery call. From there, I scope the problem, define deliverables, and propose either a fixed-scope project or a retainer. Minimum meaningful engagement is usually 40 hours.
Do you work with non-technical teams?
Yes. Most of my clients are domain experts, not developers. I handle the technical architecture and implementation. You provide the operational context — that's the most valuable part.
What industries have you worked in?
22 years in oil & gas (Repsol, ADNOC, Moeve). AI projects across scientific research, financial operations, and engineering documentation. Blockchain and tokenization advisory across multiple sectors.
How do you handle data security?
I sign NDAs before any project starts. I work within your existing security infrastructure when required and use enterprise-grade encryption for all data transfers.
How quickly can you start?
Typically 1–2 weeks after our discovery call. For urgent projects, reach out directly and we'll figure it out.
-
Ready to talk?
30 minutes. No pitch. Just a conversation about your problem and whether I can help.