VN COSTA ANALYTICS

About me

I'm Vinícius Nunes da Costa, a Data Scientist and ML Engineer based in Recife, Brazil. I build applied AI systems where machine learning engineering, clinical research, and business impact meet.

What I do

Applied AI from research to production

My work covers the full arc from research to deployment. On the technical side, that means designing and shipping ML pipelines, fine-tuning language models, building computer vision systems, and keeping production services reliable with MLOps.

I focus on domains where the stakes are real: healthcare, logistics, and data-intensive business operations. Over four years across startups and consulting, I’ve learned that the hardest and most valuable part of applied AI is turning a working prototype into something a team can maintain, trust, and improve.

Background

Formal training and context

I hold a B.Sc. in Computer Science from CESAR School (2020–2023) and a B.A. in Business Administration from FCAP–UPE (2020–2025). I chose the dual-degree path deliberately: technology alone rarely creates durable value without understanding the organizational and economic context around it.

I’m currently completing an M.Sc. in Computer Engineering at POLI–UPE (2025–2027), where my research focuses on early anomaly detection for epidemiology time series. My experience in econometrics, computational modeling, and applied AI comes from FCAP–UPE, while at CESAR School my research centered on NLP, learn-to-rank mechanisms, semantic search, and retrieval re-ranking for automatic ICD coding. I also hold a CEFR C2 English certification (Michigan ECPE, 2023).

Computer Science

CESAR School, 2020–2023

Business Administration

FCAP–UPE, 2020–2025

M.Sc. in progress

POLI–UPE, 2025–2027

Core competencies

Technical stack

Machine Learning & AI

Supervised and unsupervised learning, deep learning, NLP and LLMs, computer vision, recommendation systems.

MLOps & Engineering

FastAPI, Docker, AWS, GCP, Azure DevOps, CI-friendly templates, Poetry, dbt.

Data & Analytics

Spark, Hadoop, ClickHouse, ETL/ELT pipelines, A/B testing, exploratory analysis.

LLM Tooling

LangChain, LangGraph, HuggingFace, Guardrails AI, Elasticsearch, RAG architectures.

Domains I care about

Where I do my best work

Healthcare AI

Clinical NLP, diagnostic support, treatment decision systems, and explainability in high-stakes settings.

MLOps & production ML

Turning AI into reliable services and systems that teams can iterate on and trust in production.

Applied research

Problems between engineering and science, where rigorous method and practical judgment both matter.

Work history

Condensed timeline

Focus Distribuidora — AI Consultant

Apr 2025 – present

Sales intelligence, data governance, churn modeling, and digital transformation in logistics.

Vitally Health — Mid-level Data Scientist

Dec 2024 – Mar 2025

Clinical decision support for hypertension and heart failure, medication triage systems, FastAPI + AWS Lambda deployment.

Pickcells — Junior Data Scientist

Mar 2024 – Nov 2024

NLP/LLM pipelines for biosignal extraction, computer vision for cancer detection, and an MLOps template (ai-dat).

Oncase — Data Science Intern

Nov 2022 – Dec 2023

Ensemble clustering, LightFM recommender systems, and Big Data pipelines with Spark.

Beyond work

What keeps me engaged

I’m curious about the theory behind what I build. I’ve been a teaching assistant for Statistics & Probability and Machine Learning at CESAR School, and I think explaining things clearly is one of the best ways to find the gaps in your own understanding.

The most interesting problems in AI are interdisciplinary by nature, so I keep one foot in engineering and one in business and research.