Juan Felipe Agudelo Rios

G

Data Scientist

Grupo Fi

Jun 2024 - Current (2 years 1 month)

• Developed and deployed credit risk models (binary classification) for rural clients across Mexico, Peru, and Colombia, identifying a high-risk 5% segment where approximately 70% had defaulted credits. • Built expansion models using observational data for Colombia and Mexico, enabling data driven market penetration strategies. • Analyzed feasibility for expanding into the Philippines and Vietnam, delivering strategic forecasts to inform executive decision-making. • Led a team of two junior data scientists, introducing best practices in Python and Git workflow.

G

Specialist in Analytics

Grupo Fi

Feb 2024 - Jun 2024 (5 months)

• Optimized routing strategies for logistics, leveraging advanced analytics to save $2M in fuel costs over six months.

B

Consultantfor Andr´ es Mu˜ nos

BANCO INTERAMERICANO DE DESARROLLO (BID)

Jan 2024 - Jan 2025 (1 year 1 month)

• Served as consultant and co-author of a paper on subnational fiscal rules in Mexico and Colombia.

Client Service Trainee

Premise Data

Sep 2023 - Feb 2024 (6 months)

• Analyzed consumer goods market trends (FMCG) using Python and R, identifying portfolio expansion opportunities. • Conducted in-depth distribution assessments in traditional FMCG markets, improving regional coverage.

Research Assistant

United Nations Development Programme (UNDP)

Feb 2022 - Nov 2022 (10 months)

• Updated LATAM economic indicator databases, enhancing cross-country comparability. • Automated Excel VBA workflows, saving 10 hours weekly.

C

Research Assistant

Centro de Estudios sobre Desarrollo Econ´ omico (CEDE)

Feb 2022 - Sep 2023 (1 year 8 months)

Ignacio – Designed Python scripts to process real estate datasets, reducing manual cleaning time by 50%. – Created analytical reports for Mercado Libre on property pricing trends.

A

Research Assistant

Armed Conflict Location and Event Data Project (ACLED)

Feb 2021 - May 2022 (1 year 4 months)

• Maintained conflict database entries for Colombia, reducing data errors by 20%.

U

Guest Lecturer — Introduction to Programming

Universidad de Los Andes, School of Management

• AI Agents: Overview of AI-assisted coding with Windsurf and agent development using CrewAI.

M

Consultant

MUTA Colombia

• Led the creation of AI agents for enhanced customer interaction and process automation.

U

Remote Consultant

Universidad de los Andes, Facultad de Administraci ´ on (Minoru Higa Ph.D.)

• Created a large language model (LLM) that classifies role descriptions in Spanish. • Predicted whether each role was performed outdoor or indoor, supporting new research initiatives with 81% of cases being correctly classified.

Juan Felipe Agudelo Rios

About

Detail

Timeline

Résumé