Ana Muñoz

Designed a 3-stage GenAI content pipeline (Gemini, Veo 3.1) that accelerated course production by 400% vs. manual workflows. Applied advanced prompt engineering (few-shot, chain-of-thought) to ensure multi-module course coherence and quality. Optimized deployment on Google Cloud Run, achieving 10x faster generation speed through parallel API call management. Led AI vision and data governance for a TikTok-style microlearning platform serving frontline workers in LATAM.

As the Head of Data Science and Artificial Intelligence at Thinkcare, I lead the strategic development and implementation of cutting-edge data science and AI initiatives. I am responsible for spearheading the growth of the data science division, setting the technical direction, and ensuring the integration of AI technologies across various projects. My role involves collaborating closely with both technical and non-technical teams to leverage data-driven insights and AI solutions that enhance operational efficiency and drive business innovation. Built an advanced duplicate detection system (RapidFuzz + LLM standardization + geolocation), improving operational efficiency for a major Mexican retailer. Designed route optimization using constrained KMeans + Google Maps APIs, boosting field operations by 30%. Developed a LangChain multi-agent system for automated medical scheduling, integrating Google Calendar, Gmail & Drive APIs. Led GenAI adoption training for executives, developers and PMs — reducing dev time by 20% and timeline deviations by 15%. Managed Google Cloud partner relationship and a team of 5, delivering multiple client projects on time.

Developed custom language models on GCP to query BigQuery and deliver real-time insights in graphical, tabular, and descriptive formats. Led weekly demand forecasting for wholesale clubs (SAMS, Costco, City Club) using time series models built from scratch. Implemented forecasting for retail chains (Moyo, Chilim Balam) end-to-end: EDA, data cleaning, model comparison, and AutoML on GCP.

I did an internship with Green Spin to do my master thesis in data science. I worked closely with them and the research center CAIRO (Wurzburg, Germany). Built CNNs in PyTorch with transfer learning to detect wheat crops from satellite imagery, surpassing previous accuracy by 10%. Reduced satellite data preprocessing time by 60%. Code became the technical foundation for the research group’s future projects.

I have been helping students to catch up in university and college in various topics as: linear algebra, calculus, probability, statistics, data science, physics and more. I really enjoy sharing my passion for science with my students.

In two different occasions I gave talks to the general public about science. The first one was about Entropy and the cosmic death of the universe. The second one, as about chaos theory. I really enjoy sharing my love and passion for science.