B.S. in Computer Science from Universidad San Francisco de Quito. First author of a codebook-and-language-model framework for predicting seizure onset from scalp EEG. Former Chief Product Officer at a health-tech startup.

16 / 22held-out seizures predicted (72.7%)
280 smedian lead time
1st authorEEG + language models paper
560+pull requests in industry

01 · Research

Teaching a language model to read the brain.

Method overview: EEG windows tokenized with a k-means codebook, then GPT- and Llama-style decoders raise a pre-onset alarm
Undergraduate thesis · 2025 – 2026 · advisor: Dr. Noel Pérez-Pérez

Seizure onset prediction from scalp EEG

I tokenize 23-channel CHB-MIT scalp EEG into a discrete vocabulary with a k-means codebook (K = 256) and train compact GPT-2 and Llama-3 style decoders with causal pre-training and fine-tuning (PyTorch, Linux GPU servers).

Evaluation follows a strict person-disjoint protocol: results are reported together with the false-alarm rate and a Poisson chance-level baseline.

16/22seizures predicted
280 smedian lead time
GPT > Llamap = 0.002
PyTorchEEGTransformersVector quantizationSequence modelling

Undergraduate researcher · Bioinformatics

Universidad San Francisco de Quito · Jan 2025 – May 2026

Developed genomic analysis pipelines and reduced execution times of research workloads on a shared server.

02 · Publications & talks

Publications.

  1. 2026 · manuscript submitted

    A. Proaño, N. Pérez-Pérez, M. Herrera, O. Camacho, D. Benítez. “A Codebook-and-Language-Model Framework for Predicting Seizure Onset from EEG Signals.” Manuscript submitted for publication, 2026.

  2. October 2026 · oral presentation

    XXXIV Jornadas en Ingeniería Eléctrica y Electrónica (JIEE), Escuela Politécnica Nacional, Quito.

Plain-language summary (Spanish) →

03 · Experience

Industry experience.

  1. Feb 2026 – Oct 2026

    SaliHub (Grupo CC)

    B2B SaaS for occupational health · Quito
    Chief Product Officer & Project ManagerJul – Oct 2026
    • Promoted within four months; owned product strategy and roadmap and coordinated a team of eight developers.
    • Translated clients’ regulatory and operational needs, including Ecuador’s personal data protection law (LOPDP), into data-handling and product specifications.
    Web Lead · Backend & QA SpecialistFeb – Jun 2026
    • Led the web frontend, contributed to backend architecture and relational data models, and implemented role-based access control end to end.
    • Owned QA before production deployments; 560+ pull requests across 10 repositories.
  2. Aug 2025 – Feb 2026

    Universidad San Francisco de Quito

    Production Team Intern

    Built data pipelines and analyses on Microsoft Fabric and automated workflows with Microsoft Power Platform.

04 · Education

Education.

2021 – Jun 2026

B.S. in Computer Science

Universidad San Francisco de Quito, Ecuador. GPA 3.45 / 4.00. Selected coursework (grade A): Machine Learning, Artificial Intelligence, Statistics & Probability, Theory of Computation, Distributed Applications, Computer Security, Capstone Project.

Feb – Jun 2025

Exchange semester

Université de Technologie de Troyes (UTT), France. Design & Analysis of Algorithms (A).

05 · Skills

Skills & languages.

Machine learning
PythonPyTorchNumPyscikit-learnTransformersSequence modellingVector quantizationExperimental design
Data & software
SQL (PostgreSQL, MySQL)ETL pipelinesAWSTypeScript / Next.jsJavaC / C++GitLinux
Languages & memberships
Spanish (native)English (B2, Duolingo 125)French (basic)IEEE Student MemberAWS Academy Cloud FoundationsGoogle Data Analytics

06 · Contact

Let’s talk.

Full CV available on request.

andres.proano.dev@gmail.com LinkedIn GitHub