CV
Astronomer turned AI & data science lead — a PhD in exoplanets, now building LLM and ML systems in industry.
Professional Summary
Astronomer turned data scientist. I earned my PhD building Bayesian methods to detect and characterize exoplanets, and now lead AI and data teams in industry — shipping LLM/RAG systems, forecasting models, and the platforms that serve them.
Experience
Futuro del Trabajo
AI strategy and technical roadmap for a data team.
- Lead a team of two, defining AI strategy and the technical roadmap.
- Taxonomy-constrained text classification aligned to the CIUO occupational framework, using LLMs, RAG, and supervised ML.
- Embedding-based retrieval pipelines for structured occupational inference and semantic similarity matching.
- LLM evaluation harnesses — multi-judge panels with log-probability-calibrated scoring and hierarchical-cascade error analysis under severe class imbalance.
Toteat
First data hire; built and led the company’s data function.
- Established the data function as the first dedicated data hire, and led a junior data scientist.
- Designed ETL pipelines integrating transactional, operational, and client datasets.
- Introduced workflow orchestration with Prefect for reproducible, scheduled processes — moving analytics from ad-hoc reporting to structured pipelines.
Built the BI and predictive-modeling platform.
- Designed and maintained Toteat Analytics, the company’s BI and predictive-modeling platform on Apache Superset.
- Built churn-prediction pipelines and sales/demand forecasting models for inventory and operational planning.
- Designed NLP classification systems; deployed ML services via containerized Flask APIs and Streamlit interfaces.
MDEyeCare
CNN medical imaging for early ocular cancer detection.
- Designed and trained a CNN-based medical image classifier for early ocular cancer detection.
- Developed image preprocessing and brightness-normalization algorithms for robustness under variable acquisition conditions.
Deloitte
IFRS9 / credit-risk model validation for Chilean banks.
- Validated IFRS9 / Expected Credit Loss models for Chilean banks under Basel III requirements.
- Designed automated regulatory reporting pipelines, and built a RAROC calculator with a GUI for internal capital allocation.
Freelance
Radiotherapy treatment-plan visualization tool.
- Developed a closed-source Python application for visualizing radiotherapy treatment plans, for educational use.
Education
-
2018 - 2023 Santiago, Chile
PhD
Universidad de Chile
Astronomy
- Thesis: exoplanet detection and characterization using Bayesian formalisms.
- Advisor: Prof. James Jenkins.
-
2014 - 2018 Santiago, Chile
BSc
Universidad de Chile
Astronomy
- Graduated cum laude.
Awards
-
2022 Start Up Ciencia SUC 220075 — CTO
ANID (Chilean Government)
-
2019 National Doctorate Scholarship
ANID (former CONICYT)
One of 734 doctoral scholarships awarded in Chile that year.