CV

Download PDF

Vinícius Nunes V. da Costa

Data Scientist & ML Engineer · Recife, Brazil (Remote) LinkedIn · GitHub · viniciusnvcosta47@gmail.com

Data Scientist and ML Engineer with 4+ years of experience building and deploying AI solutions in startups and healthcare. Strong background in Python, machine learning, deep learning and MLOps on AWS and other major clouds. Experience delivering NLP/LLM, computer vision and recommender systems from prototype to production, partnering closely with product and business teams.

Experience

Focus Distribuidora — AI Consultant

Apr 2025 – Present

• Managing and developing sales intelligence initiatives, enhancing data-driven decision-making across the organization.
• Establishing a robust data governance framework, fostering a culture of analytics and accountability.
• Applying ML for churn analysis, leading to improved customer retention strategies.
• Supervising digital transformation projects in logistics, optimizing operational efficiency and reducing costs.

Vitally Health — Mid-level Data Scientist

Dec 2024 – Mar 2025

• Designed a triage system for hypertension and heart failure (HFrEF/HFpEF) that ranked medication titrations and prescription changes using patient-specific clinical constraints and AHA guidelines.
• Developed categorical encodings for comorbidities, vitals, labs, and contraindications to enable accurate and explainable clinical decision-making.
• Converted unstructured medical guidelines into structured knowledge bases and business rules.
• Built and deployed model-as-a-service endpoints using FastAPI, Pydantic, and AWS Lambda.
• Implemented complementary modeling strategies combining rules and models to balance personalization, safety, and guideline compliance.

Pickcells — Junior Data Scientist

Mar 2024 – Nov 2024

• Conducted clinical research and developed AI products for diagnostic support in Brazilian healthcare.
• Built an NLP/LLM pipeline to extract health scores from biosignal data directly from PDF blood work exams.
• Contributed to computer vision models for cervical cancer detection (40× microscope imagery) and acute leukemia detection in children's myelograms.
• Led MLOps initiatives, creating a Python cookie-cutter AI deployment template (ai-dat) with FastAPI, Docker and tests, reducing deployment time by ~50%.
• Co-led the data science team as a technical reference, mentoring on best practices in AI deployment.

Oncase — Data Science Intern

Nov 2022 – Dec 2023

• Designed an ensemble clustering model using Autoencoder + KMeans, improving customer segmentation quality by ~20%.
• Refactored a legacy B2B recommender system using LightFM for hybrid recommendations, optimizing product mix and pricing strategies.
• Created data visualization dashboards with Metabase and Python for consulting clients.
• Managed Big Data pipelines with Spark and SQL on cloud platforms.

Education

M.Sc. Computer Engineering (in progress)

POLI – UPE · 2025 – Expected 2027

Research focus: econometrics, computational modeling, and applied AI. Research grant through PROPEGI.

B.Sc. Computer Science

CESAR School · 2020 – 2023

Thesis: Automated ICD-10 Coding Suggestion in Brazilian Electronic Health Records: a deep learning approach using Transformer networks and NLP. Teaching assistant for Statistics & Probability (2022) and Machine Learning (2023).

B.A. Business Administration (dual degree)

FCAP – UPE · 2020 – 2025

Certifications

• Deep Learning Specialization — DeepLearning.AI (2025)
• Generative AI with Large Language Models — DeepLearning.AI & AWS (2024)
• Michigan ECPE (CEFR C2 English) — University of Michigan (2023)

Skills

Programming & Data Science — Python, SQL, Pandas, NumPy, Scikit-learn, TensorFlow, PyTorch, Ultralytics, OpenCV

Machine Learning & AI — Supervised/Unsupervised Learning, Statistical Learning, Deep Learning, NLP, LLMs, Recommendation Systems

Cloud & MLOps — AWS (S3, EC2, SageMaker, Bedrock, Lambda, Athena, RDS, EMR), GCP (Compute Engine, Vertex AI), Azure (DevOps, VMs); Docker, FastAPI, Poetry, dbt, uv, Git

Data & Analytics — Spark, Hadoop, ClickHouse, ETL/ELT, A/B Testing, Metabase, MySQL, Oracle

LLM Tooling — LangChain, LangGraph, HuggingFace, Guardrails AI, Elasticsearch, RAG

Languages

• Portuguese — Native
• English — Proficient (CEFR C2, Michigan ECPE – 2023)
• Spanish — Intermediate

Selected Projects