available for select work

I build systems
that learn.

about

Engineer and technical leader with 9+ years building data systems, optimization models, and decision-support tools in the energy sector. Now co-founding Fikko — a fintech for shared financial management — where I own the full stack and product decisions end-to-end.

experience

Co-founder & CTO · Fikko

2025 — now

Designing and building the full product stack: Next.js, Django, Postgres (Neon), Redis, Celery — deployed on Railway.
Built an ML-based automatic transaction categorization model integrated into the core product.
Driving product, engineering, and business decisions end-to-end as the sole technical co-founder.

Head of Digitalization · Statkraft

2023 — 2025

Led the digitalization team for Statkraft's energy trading arm in Brazil.
Designed and oversaw revenue projection systems (Synapse + PySpark) and BI infrastructure used daily by trading and risk teams.
Managed analysts, set technical roadmap, and aligned with trading, risk, and executive stakeholders.

Senior BI Analyst → BI Specialist · Statkraft

2020 — 2023

Developed credit risk and market risk evaluation application.
Implemented a new ETRM system and integrated it with the risk application via APIs and messaging.
Built a Datalake and ETL pipelines on Azure Synapse; maintained Power BI dashboards used across the business.

Software Engineer & Quantitative Developer · PSR Energy Consulting

2017 — 2020

Implemented stochastic optimization models for energy planning (SDDP model contributions).
Built and maintained the full CI/CD infrastructure and automated testing platform for the engineering team.
Delivered models for international clients: Colombia's energy regulator (CREG) and Peru's Ministry of Mines and Energy.

education

Postgraduate Specialization in Software Engineering · PUC-Rio

mar 2025 — dez 2026

Software architecture (microservices, cloud, DevOps), agile project and product management, and software quality and testing.
Intelligent software systems: ML model design, deployment, and automated testing.
480-hour program structured in sprints, each ending with a practical MVP project.

Professional Certificate in Finance · COPPEAD UFRJ

out 2022 — jun 2023

Corporate finance, capital structure, financial planning, and company valuation (DCF).
Studied derivatives, fixed income, portfolio management, risk management, and ESG frameworks.
160-hour program with academic supervision by COPPEAD/UFRJ faculty, one of Brazil's top-ranked business schools.

Master of Engineering – Decision Support Methods · PUC-Rio

2018 — 2020

Coursework: stochastic processes, statistical inference, multivariate statistical methods, linear and nonlinear programming.
Applied optimization and probabilistic modeling within electrical engineering and power systems contexts.
GPA: 9.1/10.

Master of Engineering – Automated Systems · CentraleSupélec

2014 — 2016

Double degree with Escola Politécnica / UFRJ under the international partnership agreement.
Deep focus on control theory: state-space, predictive, nonlinear, and hybrid systems control; estimation, filtering, and identification.
Coursework also covered power conversion, energy production, numerical methods, and optimization.

Eletrical Engineering · Escola Politécnica da UFRJ

2011 — 2017

stack

PythonPySparkNumPyPandasSQLKafkaAzureDeltalakeSynapsePower BIMetabaseDevOpsMachine LearningOptimizationStatisticsData EngineeringPostgresRedisDockerJuliaC++TypeScriptNextJSDjangoCeleryJenkinsJavaScript

selected work

Simplex Benchmark

Benchmark of the two-phase Revised Simplex in Rust, Python, Julia, C++, Go, and TypeScript. Netlib LP instances; from-scratch pivot logic, timed on median wall-clock and peak RSS.

JuliaPythonC++GoRustTypeScript

Hourly Energy Contract Pricing

Pricing engine for hourly energy contracts, using a stochastic programming model to optimize the contract price. The model is a mixed-integer linear program (MILP) that is solved using the Gurobi solver.

JuliaPythonGurobiPyomo

Energy Price and Consumption Forecasting

Forecasting energy price and consumption using a time series model. The model is a linear regression model that is solved using the PyTorch library.

PythonPyTorchPandasNumPy

writing

The start of Fikko

2026.06.01 · 3 min read

Why I left a confortable job in the energy industry.

contact

Simplex Benchmark

Hourly Energy Contract Pricing

Energy Price and Consumption Forecasting

Let's build something