Juan Lara
Computer Scientist Applied Mathematician Machine Learning AI Data Science

Juan Lara

Research Assistant at Harvard Business School

"The beauty of mathematics only shows itself to more patient followers."

— Maryam Mirzakhani
Download my CV

About Me

I am a Computer Scientist and Mathematician trained at the Universidad Nacional de Colombia, with specialized expertise in Machine Learning. Currently collaborating on applied research at Harvard Business School with Professor Jorge Tamayo, where I merge theoretical rigor with computational solutions to study the structure of firms.

Throughout my career, I have developed advanced mathematical models, implemented numerical simulations, and applied machine learning techniques to transform data into strategic insights. My interdisciplinary approach enables me to turn complex theories into practical and efficient tools that drive real-world impact.

I am particularly interested in AI Agents & Multi-agent Systems, exploring the development and optimization of autonomous AI agents and their interactions within multi-agent environments. My research also focuses on Computational Organizational Theory, investigating mathematical models of organizational structure, knowledge flows, and decision hierarchies using computational simulations and empirical validation.

Experience

Research Assistant

Harvard Business School

September 2022 - Present

Conduct advanced research in organizational economics with Professor Jorge Tamayo, focusing on two key areas:

  • Computational Implementation — Translated theoretical concepts into actionable insights by developing ML models (XGBoost, clustering), implementing NLP pipelines for technical document analysis, and building modular data architectures that integrate statistical analysis with visualization techniques (UMAP, SVD).
  • Theoretical Modeling — Formalized organizational hierarchies and resource optimization through economic frameworks based on established literature (Garicano, Caliendo & Rossi-Hansberg), creating rigorous mathematical proofs to analyze policy impacts on productivity.
Machine Learning NLP Data Architecture Visualization XGBoost

Data Scientist

Ipsos

February 2024 - January 2025

Transformed market research operations through strategic data science implementations:

  • Field Operations Innovation — Developed a mobile R Shiny application with embedded machine learning models for pharmacy segmentation, integrating geospatial mapping (Leaflet) with cloud storage, resulting in 85% reduction in manual classification work.
  • NLP-Powered Analytics — Created TextInsight, a comprehensive Python library using transformer models (BERT) and network visualization (NetworkX) for multilingual survey analysis, reducing processing time by 60% while delivering more nuanced insights.
  • Interactive Reporting — Designed dynamic dashboards and visualization tools that streamlined decision-making processes across Latin American operations, earning recognition as Total Ops Star Employee for LATAM.
R Shiny Leaflet BERT NetworkX Cloud Storage

Education

B.S. in Mathematics

Universidad Nacional de Colombia

2018 - 2022

GPA: 4.7/5.0 Applied Mathematics
  • Focus on Mathematical Analysis and Optimization
  • Research in Game Theory and Economic Modeling
  • Advanced courses in Numerical Methods

B.S. in Computer Science

Universidad Nacional de Colombia

2019 - 2023

GPA: 4.7/5.0 Machine Learning
  • Specialization in AI and Machine Learning
  • Advanced Algorithm Design and Analysis
  • Software Engineering and Development

Additional Training

Featured Projects

A modern, LaTeX-enabled static website for mathematics and computer science notes. Features category-based organization, beautiful mathematical typography, and responsive design.

Python Flask LaTeX KaTeX

TextInsight - Python Library

Multilingual text analysis library combining BERT for sentiment analysis, GPT-3.5 for text correction and topic generation, and embeddings for graph visualization. Deployed at Ipsos to analyze survey data, cutting analysis time by 60% while delivering deeper insights.

Python Transformers OpenAI NetworkX PyVis

Pharmacy Segmentation Application

Responsive mobile field application for pharmacy segmentation using R Shiny with Google Cloud Storage integration. Features geolocation mapping with Leaflet, route management, and real-time ML classification, reducing manual segmentation effort by 85%.

R Shiny Google Cloud Random Forest Leaflet

Cunservicios Platform

Web platform for water and sewage utility services management with a modern microservices architecture. Features user modules for bill consultation and payment, service requests, and PQR management with responsive design.

React FastAPI SQLAlchemy Tailwind CSS

Multi-Agent Optimization

Research project on optimization in organizational economics with Harvard Business School, focusing on game theory applications.

Game Theory Python Optimization

Web-based game developed using JavaScript with advanced gameplay mechanics.

JavaScript HTML5 CSS3

Let's Connect

Feel free to reach out for research collaborations, technical discussions, or just to say hello!