I build LLM-powered systems end-to-end — and measure them with a statistician's rigour.
- 🤖 AI Engineer — I design and ship LLM systems end-to-end: prompt pipelines, LLM-as-judge evaluation, APIs and self-hosted GPU apps.
- 📊 Foundation in Statistics & Data Science — professional experience as Data Scientist and Statistician across consulting, market research, scientific research and the consumer-goods industry.
- 💻 Completing a B.Sc. in Information Technology (UNIVESP, 2027) to formalize the software-engineering side.
- 🧪 What that combination brings — AI systems that are measured, not guessed: sound evaluation, quantified trade-offs, reproducible and well-documented decisions.
LLM evaluation & responsible AI
| Project | Summary | Stack |
|---|---|---|
| llm-eval-stats | Python library for statistically sound LLM evals: bootstrap CIs, paired model comparisons (McNemar/permutation), LLM-judge vs human agreement and sample-size planning. | Python · NumPy · SciPy |
| ai-generated-content-evaluator | Generates technical reports with NotebookLM and auto-evaluates them with Gemini across four quality metrics. | Python · Gemini API |
| apertus-ethics-by-design-case-study | Maps the Swiss Apertus LLM to the EU Ethics by Design framework and AI Act, quantifying the compliance/performance trade-off. | AI governance · LaTeX |
| ai-fluency-ptbr | Brazilian-Portuguese translation of A Framework for AI Fluency with an interactive companion. Live demo → | Tailwind · Chart.js |
Applied AI & self-hosted systems
| Project | Summary | Stack |
|---|---|---|
| whisper-transcriber | Self-hosted web app for GPU audio/video transcription — resumable uploads, real-time progress, 5 export formats. | SvelteKit · FastAPI · Redis/ARQ · faster-whisper · Docker |
| literature-reviewer | Turns a title + abstract into a structured literature review: research questions, top papers (IEEE-formatted) and speaker questions. | Python · OpenAI · Semantic Scholar |
Also: DeepSeek · Open WebUI · AnythingLLM · Hermes Agent · Claude Code & Codex CLI · LaTeX · Debian/Ubuntu home lab
- 🎓 M.Sc. in Statistical Modeling — UNICAMP / FEA · PLS regression applied to Sensory & Consumer Science
- 🎓 B.Sc. in Statistics — UNICAMP
- 🔬 PhD-level coursework (special student) — UNICAMP / FEEC · Responsible & Ethical AI · Seminars in Computer Engineering
Certifications
- Machine Learning Specialization — DeepLearning.AI / Stanford Online (2025)
- 5-Day Gen AI Intensive Course — Google × Kaggle (2025)
- Google Data Analytics Professional Certificate — Google (2024)
- Statistical Learning, with Distinction — Stanford Online (2020)
- Data Science Specialization — Johns Hopkins University (2017)
Open to roles and collaborations in AI Engineering · Applied AI · LLM Evaluation · AI Governance · AI Data Science & Analytics — and always happy to chat about LLM systems, AI agents and evaluation.