Software Developer · Machine Learning Engineer
Hermosillo, México · open to remote ML & data roles

Full-stack developer at NeuralGT, where I build Fintrack, a personal finance SaaS for the Chilean market (FastAPI + PostgreSQL + Next.js).

On the side, I am transitioning into Machine Learning Engineering: shipping production-grade ML projects, studying graduate-level statistics and optimization, and aiming for a remote ML role by late 2026.

Recent CS graduate (UTH 2026, 97/100 GPA).

Geo-contextual player segmentation & content ranking

Production-grade ML pipeline simulating a Pokémon GO–style game with 50,000 synthetic players, 162M events, and 5 behavioral archetypes (commuter, casual evening, weekend explorer, hardcore raider, lunch player). The pipeline clusters players via HDBSCAN on a 54-feature matrix (temporal density, H3 spatial footprint, session behavior) and ranks content with LightGBM.

Highlights

• Partition-streaming feature pipeline bounded to ~3 GB constant memory on 162M events
• 54 engineered features across three families: temporal (39), spatial with H3 (7), behavioral (8)
• Strict tooling: uv, ruff, mypy, pre-commit hooks
• MLflow experiment tracking, FastAPI serving layer, Dockerized deployment, GitHub Actions CI/CD

Full-stack SaaS for personal finance in Chile. FastAPI backend with Alembic migrations, PostgreSQL on AWS RDS, Next.js frontend with Feature-Sliced Design. Integrations with Floid (banking sync), Binance, and mindicador.cl.

Languages

Machine Learning & Data

Backend

Frontend

Tooling

• Shipping geoplay-recommender end-to-end: clustering → ranking → MLflow → FastAPI → Docker
• Strengthening statistics and optimization fundamentals for graduate-level ML work
• Building fluent technical English for international remote collaboration

When I'm not coding, I write literary fiction (currently a novella anchored in Camus and Kierkegaard), explore FromSoftware games, and overthink internet culture.

A small Yanma to close. Field guide entry: Bug/Flying type, Generation II, Pokédex #193. Known for its compound eyes that see 360° around itself a useful trait for someone who works with data from many angles :D

                            .::::@-   .-::$-
                             ;::;;;;#;:::;;+ -H#@
 *&&-        :      :*              *;H;+++;+++++++;.
 -:&;;;+#  ,&:     ::               :+++$@    @H+++#
   ;++++++;;;+*   ;;;             --H;*#
     #     .;;+  +++             -::;**
        +   ;;*+++++#H         :;::$*  #       ;+;;+@  @
           #+;;++$::::+*:::::@+&@;,        ;@*&@@+**&$$$@
          &;+:;;::::::;**;+;+&,$- .....        ;;;;;;;     -
          ,:$;++::::;:+**@,,,,,,.             ;+++;;;+
         H:+++++;::H #***&***@*+.-+*+H#H@HH&*$$#$$$#;
         @+++++++++$#&***$*H**# ;#@@@
          *@;;:*+**$@$****H*+@@$   $@$
           #+**H***@@@**@:@    &@   @@
           ;***&*@****&   @    @@   H**
             &#$&#       ,@    @@    ***
           @&#  .        @@   #@     *
         ***            @@    **
           H          #@@     +@
                     **+
                     @#;

_{Thanks for stopping by. Reach out anytime via email or LinkedIn.}