Tech entrepreneur · Product builder · AI/Data/Fintech
I’m a Brazilian tech entrepreneur and hands-on product builder with 16+ years in software, working across AI, data, fintech and market intelligence — from real systems, not slide decks.
16+ years building software · +20k SAAS users · AI/data products · Fintech workflows · Building IrishTalents
Building now
Current work across market intelligence, fintech automation, practical AI systems and operating contexts where software has to meet reality.
Current product
A data product mapping Ireland's work-permit sponsorship market.
It structures public employment-permit data, enriches company profiles with AI and turns sponsorship signals into search, insights and market intelligence for candidates, recruiters and partners.
Explore IrishTalentsAI-assisted product and architecture work for investment banking workflows, document-heavy operations and capital-markets infrastructure.
The focus: making complex financial processes more structured, searchable, automated and actionable.
View fintech workPractical experiments with LLMs, RAG, agents, pgvector, LangGraph, CrewAI, Ollama, scrapers and workflow automation.
Focused on what can become a real product surface, not AI theatre.
See experimentsDigital product exploration around transport operations, payment checking and financial control.
A practical operating context where software, automation and real-world business constraints meet.
View project notesProof from shipped work
A compact view of the proof: senior software work, startup execution, real users, paying customers and the product I’m building now.
How I work now
About
Now
Collaborate
Technical proof
The stack matters. But the real advantage is knowing when to code, when to simplify, when to ship, and when architecture is becoming a business bottleneck.
Leading small teams, shaping product decisions and moving between strategy, code and delivery when momentum matters.
Still close enough to the code to understand trade-offs, constraints and what can actually ship.
Search, enrichment, RAG, agents, document workflows and data products built around real operational needs.
Cloud, deployment, observability, cost control and systems that can survive production.
Applied AI
A short view of where I’m applying AI now: document-heavy workflows, agent orchestration and searchable data products.
Classification, extraction, RAG and reporting workflows applied to finance, IBaaS and document-heavy operations.
LangGraph, CrewAI, worker queues and scoped agent systems tested with human review around risky actions.
pgvector, embeddings, enrichment and market-intelligence products built from public and operational source data.
Journal
I write field notes about building products, using AI pragmatically, creating data assets, exploring European markets and learning from the products I’m building in real time.
This is not a content funnel. It is a public notebook for decisions, lessons, trade-offs and systems.
Why useful background agents need triggers, queues, reviewable outputs, and clear human ownership more than they need vague autonomy.
Why agents need process, training material, and verification steps before their work becomes predictable enough to trust.
How I think about connecting Search Console, PostHog, and Hermes into a weekly review loop for acquisition, product behavior, and follow-up work.