CV

Education

• M.S. in Mathematical Modeling, Research, Statistics, and Computation, University of the Basque Country
• M.S. in Applied Mathematics, Santo Domingo Institute of Technology
• B.S. in International Business & Finance, Utah State University

Work experience

• 2026-present: Senior AI Engineer
  • Pulsecity
  • Key contributions:
    • Designed the architecture of an AI event intelligence platform on Python and FastAPI — async ingestion workers, a Postgres-backed canonical event schema, and a service boundary that downstream products read from.
    • Built an LLM enrichment pipeline where swappable agents (OpenAI, Anthropic, Llama) enrich raw events through a Model Context Protocol layer with schema validation and confidence scoring, so the stack can follow model progress without a rewrite.
    • Shipped a configuration-driven ingestion framework that onboards any event source — REST, GraphQL, or scraper — under one adapter contract, with deduplication and geo-resolution built in.
    • Wired observability into every stage (structured logs, per-agent audit trails, quality scoring) so data drift and enrichment regressions are measurable and surface before they reach users.
• 2025-2025: Senior ML/Software Engineer, Pricing
  • HP Inc (via Mphasis)
  • Key contributions:
    • Collaborated within the Pricing Data Science team of HP Inc (Sant Cugat), with a data science software refactoring project for various markets, aiming at developing a common architecture for feature engineering, modeling, and deployment of price elasticity and causal inference models.
    • Architected 3 modular Delta Live Tables pipelines in Python to support retail and commercial demand models across North America and EMEA, reducing code duplication by ~11%, enabling region-specific customizations.
    • Built counterfactual price model (LightGBM) to simulate competitor brand prices as if they were an HP brand, to use as confounders in a demand elasticity model.
• 2024-2025: LLM Engineer, Post-training
  • Invisible Technologies
  • Key contributions:
    • Executed advanced post-training workflows, including Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF), to optimize model reasoning and alignment.
    • Developed complex evaluation datasets including Chain-of-Thought (CoT) mathematical reasoning prompts and benchmarks to measure model performance in specialized domains, ensuring high-fidelity outputs for frontier LLM providers.
    • Collaborated on the implementation of reward modeling and preference-based optimization (DPO/PPO) to refine model behavior, safety guardrails, and surface model failure modes in complex tasks.
• 2023-2024: Data Scientist, Promotion Optimization (BEES)
  • AB InBev
  • Key contributions:
    • Collaborated on a promotion (pricing) optimization algorithm, improving key metrics like ROI, investment, and coverage across various promotional strategies (combos, stepped, among others).
      • This involved modeling demand elasticity with a log-log model (XGBoost), optimizing for a chosen metric with cubic splines, rank suggested order arrangements in combos, among other dynamic phases.
    • Optimized ROI by 48% in A/B test promotional experiments, over a period of 8 months.
    • Developed an algorithm targeting first-time purchasers, optimizing discount allocation and saving on budget.
    • Created an RCT module that automated user allocation into control and treatment groups and logged experiment metadata (promotion ID, allocation, blocking factors, timestamps) to a historical registry, reducing experiment design time by ~80%.
    • Implemented uplift and causal ML models (X-Learner, DR-Learner, Synthetic DiD) to measure market innovation effectiveness (ATT, CATE).
• 2022-2023: Data Scientist, Product Recommendation
  • Santa Cruz Bank
  • Key contributions:
    • Optimized the product recommendation (NBA) model architecture, enhancing feature engineering and embedding calculations, into a batch, two-tower modeling approach.
    • Enhanced customer segmentation by incorporating digital behavior clustering and psychographic profiling based on commerce data.
    • Automated weekly client account reporting via a data pipeline job, boosting productivity 2x and eliminating manual data wrangling.
    • Applied NLP techniques (sentiment analysis, n-grams, entity classification) to extract insights from Salesforce CRM interaction data.
• 2019-2022: Data Scientist, Risk Analytics
  • DGII
  • Key contributions:
    • Collaborated in developing the analytics & machine learning feedback system to label taxpayers as risky or not risky.
    • Built models such as linear regression, decision trees, PCA, logistic regression, MCMC, and clustering.
    • Helped define sector-specific, multidimensional risk metrics spanning income underreporting, cost inflation, shareholder benefit abuse, and transfer pricing anomalies.
    • Mapped shareholding structures with graph theory (NetworkX, Bokeh), leveraging centrality insights to reveal influential entities and hidden transactional links, improving tax evasion case prioritization by 2x.
    • Created the Data Warehouse inventory eCatalog, a Shiny (R) solution mapping all data infrastructure used for risk estimations (metadata, data dictionaries, data lineage). This streamlined the process of identifying data sources by 2x.
• 2017-2019: Junior Data Analyst, Tax
  • Deloitte
  • Key contributions:
    • Supported tax compliance analytics and client engagements through requirements gathering, data modeling, and reporting.
    • Built workflows to consolidate financial and transactional datasets for audit and tax risk reporting.
    • Collaborated with cross-functional consulting teams (Finance, Transfer Pricing, Audit, Legal) in financial projects, including M&A and tax risk analyses, to deliver data-driven insights to clients.

Skills

• Python
• R
• SQL
• Git
• Spark
• Data Science & AI Techniques: Propensity score matching (PSM), Double machine learning (DML), Doubly robust machine learning (DRML), A/B/n testing, RCTs, CATE calibration methods, RAG, among others.
• Azure Cloud: Azure Databricks, DL/Hive, Azure Machine Learning Studio, Azure DevOps, Unity Catalog, Lakeflow (data orchestration), Delta Live Tables (DLT), and Azure Synapse Analytics. AWS: Amazon SageMaker, Amazon Redshift, Athena, AWS Glue.
• Additional Tools: MATLAB, Java, Gurobi, PuLP, Docker, MLflow, Airflow, Pytest, LangChain, MCP, CI/CD.

Leadership

• AB InBev Analytics Workshop Professor (2024)
• AB InBev Global Hackathon (2023)
• R Programming Mentor (2022-2023)

Last updated: 2026-04-18