Experienced engineer. Applied AI focus.

Trajectory

1. 2018 — 2021

   Data Analyst · Arbusta · Buenos Aires

   Started in tech on the data side — queries, dashboards, and the discipline to chase root causes through real datasets. The foundation for how I think about evaluation and instrumentation now.

2. 2021 — 2023

   Web Developer · Arbusta · Argentina (Hybrid)

   Moved from analytics into engineering. Built internal tools, dashboards, and client-facing interfaces on Scrum teams. GitHub PR reviews, CI-backed releases, real product work.

3. 2023 — 2024

   Frontend Developer · Huenei IT Services · Remote

   React on a distributed Agile team. Modular components, REST integration, and performance tuning — lazy loading, code splitting, render optimization — to support growing concurrent load.

4. 2024 — Present

   Software Engineer · Suncor Stainless · Plymouth, MA (Hybrid)

   Full-stack ownership of a production web platform. Led a mobile-first redesign with Tailwind that lifted Core Web Vitals 25%+. Architect hybrid React/Vite front-ends that bridge a legacy PHP/SQL backend — incremental modernization without service interruption. Build automation that replaces manual operational work, and use AI-assisted tooling daily as part of the development workflow.

5. 2026

   Applied AI, in production · Open source

   Building extraction, retrieval, and agent systems on the Anthropic API. Quantara live; regulatory-intelligence RAG shipping; personal-knowledge MCP server next; workflow-automation agent on deck.

How I work

• Direct API, not framework wrappers

  I work against the Anthropic API in primitives — tool use, schemas, vision fallback, prompt caching. Frameworks abstract the parts I most need to understand. I learn them in a week when a job requires one.

• Hybrid systems, not rewrites

  Years of production work mean I know modernization happens incrementally. At Suncor I bridge a legacy PHP/SQL backend with modern React + Vite without service interruption. Same instinct applies to AI: ship the smallest useful system, evaluate, expand.

• Evals before scale

  Golden datasets, regression checks, and honest failure reports are non-negotiable. A demo that works once isn’t shipped. Every project comes with the harness that proves it still works tomorrow.

Education & certifications

• BS, Applied Technology — Information Technology

  Brigham Young University–Idaho

  2022 — 2024

• Certified Tech Developer

  Digital House

  2022 — 2023

• Front-End System Design · Next.js v3 · PHP

  Frontend Masters

  2024 — 2025