๐Ÿ‘จ๐Ÿปโ€๐Ÿ’ป My work

What I have been doing lately and what I have done in the past

๐Ÿ’ผ Current Roles

Data Scientist @ AtlasIntel

AtlasIntel provides advanced data products for political intelligence and economic forecasting.

Researcher @ MAPE

MAPE is dedicated to studying public policies, elections, and their connections. Our goal is to improve academic production on the subject and to have a decisive impact on public debate. In particular, we combine different methodologies to study policies and elections, including causal inference techniques and machine learning.

If you think I can help you in any way, please reach out at felipe.lamarca@hotmail.com.


๐Ÿ“ Past Work

๐Ÿ“ Consultant in Data Science & Research (07/2024 โ€“ 06/2025)

Provided end-to-end data and AI solutions (ETL, ML, Generative AI, RAG), research support, and training.

Selected consulting projects:

  • ๐Ÿ“Œ Artplan: Deep Learning for social media segmentation
  • ๐Ÿ“Œ Fundaรงรฃo Roberto Marinho: AWS data lake, dashboards (Streamlit and Power BI), Python/ML training
  • ๐Ÿ“Œ CEBRAP: OCR, LLMs, and quantitative research support for Prof. Josรฉ Szwako
  • ๐Ÿ“Œ Plataforma CIPร“: Training in ML and data analysis
  • ๐Ÿ“Œ GENI/UFF: Built LLM and RAG-based data tools

๐Ÿ“ Data Scientist & Researcher @ Instituto Rio21 (04/2022 โ€“ 06/2025)

I conducted quantitative research such as opinion polls and other surveys such as the [Municipal Management Assessment Survey] and [Election Polls]. I also collaborated with Alianรงa CentroRio to transform citywide conservation and security data into actionable intelligence.

๐Ÿ“ Research Assistant @ FGV EBAPE (02/2025 โ€“ 04/2025)

I worked on a project to systematize databases for election results from the 1970s, using OCR and large language models to digitize and format historical documents. I worked under the supervision of Prof. Cesar Zucco, Ph.D.

๐Ÿ“ Data Science Intern @ Strategic Planning Superintendence, FGV (10/2024 โ€“ 12/2024)

I integrated advanced Generative AI techniques into institutional data pipelines, identifying opportunities for automation and process improvement. I also prepared presentations for C-level executives, including the President of FGV.

๐Ÿ“ Part-time Data Scientist @ Visagio (01/2024 โ€“ 06/2024)

I worked on optimization, pricing, A/B testing, and data pipeline automation with SQL. I developed Power BI dashboards and presented KPI-driven insights to senior management.

๐Ÿ“ Student Consultant @ Brazilian Sailing Confederation (08/2023 โ€“ 12/2023)

I collaborated with colleagues to develop a ranking model for Brazilian sailors, integrating web scraping and machine learning to analyze regatta performance.

๐Ÿ“ Junior Research Fellow @ FGV CPDOC (06/2019 โ€“ 09/2022)

Supported by an FGV/CNPq scholarship under the supervision of Prof. Jaqueline Porto Zulini, Ph.D., I examined Executive-Legislative relations in the early 20th century, applying both qualitative methods and introductory quantitative text analysis.


๐Ÿ“š Teaching Assistance

Semester Course Institution Professor
2024.2 Quantitative Methods II FGV CPDOC Jairo Nicolau, Ph.D.
2020.2 Introduction to R FGV CPDOC Jimmy Medeiros, Ph.D.

(Previously, I also worked on โ€œDigital Literacyโ€ projects and translated materials for the Programming Historian in Portuguese in partnership with Universidade Nova de Lisboa.)


๐Ÿ“ž References

Name Institution Email
Prof. Jaqueline Zulini, PhD School of Social Sciences (FGV CPDOC) jaqueline dot zulini at fgv dot br
Prof. Luiz Max Carvalho, PhD School of Applied Mathematics (FGV EMAp) luiz dot fagundes at fgv dot br
Prof. Renato Rocha Souza, PhD University of Vienna renato dot rocha dot souza at univie dot ac dot at
Prof. Danielle Sanches, PhD School of Communication, Media and Information (FGV ECMI) danielle dot sanches at fgv dot br
Prof. Jimmy Medeiros, PhD School of Social Sciences (FGV CPDOC) jimmy dot medeiros at fgv dot br