Loading
Loading
I'm a software engineer from Puerto Rico, in my final semester. The domains have varied a lot — radio telescope instrumentation, an SMS marketing platform, esports analytics, threat intelligence — and what I specialize in now is the part that carries across all of them: the architecture around a system's clever bit, rather than the clever bit itself. Increasingly that clever bit is a model.
Integrating AI well is mostly ordinary engineering done carefully. On ARGUS the language model orchestrates and phrases, but never sources a fact — every claim in a briefing traces back to a cited API response, by construction. On the academic platform a data-residency constraint pushed the design toward local inference and a single database holding both the vectors and the relational data. In both cases the model is one component with a failure mode, designed for like any other dependency that can be wrong.
Before that: full-stack product work at LeadWire and Red Ventures, and research engineering at the Arecibo Observatory, where I built instrumentation software for radio science. Right now I'm finishing my degree and learning Rust by writing a shell.
Jun — Aug 2026
Evertec
Built ARGUS, an AI-assisted OSINT platform that enriches indicators with risk scores, relationship graphs and MITRE ATT&CK-mapped intelligence across 14 sources. Designed a deterministic source-verification layer that cites API-derived facts independently of the LLM, so every claim in a generated briefing is auditable.
May 2024 — Aug 2025
Led a UI redesign of a large-scale SMS marketing platform — dashboard, navigation, modals and global styles — and shipped real-time dashboard updates over WebSockets. Drove a full-stack upgrade to Node 20, React 18, Apollo v3 and MUI v5, and migrated the build to Vite.
May — Aug 2023
Delivered front-end and back-end features across several sites with the Platea team. Built a 'Save' feature for blog content and an API for user-generated lists, and coded two content gates to drive A/B testing on conversion pathways.
May — Aug 2022
Arecibo Observatory
Independently built a proof-of-concept MERN application surfacing real-time data from the 12m radio telescope, with a logging tool for operational data collection.
Sep 2021 — May 2022
SEA UPRM
Developed web modules in PHP/Laravel/MySQL for the organization's services, gathering requirements with multidisciplinary teams and supporting end users.
Sep 2020 — Feb 2021
VIPPR
Processed thousands of home-reconstruction requests for an organization aiding those affected by the 2020 Puerto Rico earthquakes, extracting insights from demographic and financial datasets to inform resource allocation.