AI Engineer Jr. Data Scientist Neurocientífico

Sebastián
Mesch Henriques

Licensed Psychologist and Data Scientist bridging cognitive neuroscience and deep learning. Academic Coordinator at Humai, researcher at Universidad Favaloro, and MSc in AI student at UdeSA.

SMESCH1 · LinkedIn · Buenos Aires, open to remote

01

Featured Projects

all repos →
Neuroscience · Deep Learning · fMRI

ASD Detection via fMRI (Thesis)

Binary ASD/Control classification on rs-fMRI functional connectivity from ABIDE (1102 subjects, AAL2 atlas, 94 ROIs). Benchmarked at AUC 0.79, exploring MLP and GNN architectures.

scikit-learn PyTorch / GNN MLflow fMRI / ABIDE
NLP · Cognitive Science · Research

Ontogenia Artificial

Longitudinal study of syntactic emergence in Pythia and its alignment with human language acquisition. NLP research project investigating U-shaped learning curves.

Pythia Transformers Python Data Science
Deep Learning · Medical Imaging · CV

ABIDE DANN: Computer Vision

ASD vs control classification from rs-fMRI treated as a computer-vision problem. Uses CNNs with adversarial domain generalization (DANN) to align features across acquisition sites. M.Sc. AI project.

PyTorch CNN DANN fMRI
ML · Mental Health · Neuroscience

Autism Classification: BrainHack

KMeans clustering + Logistic Regression on neuroimaging-derived features for ASD spectrum classification. Developed during an international BrainHack workshop.

scikit-learn pandas neuroimaging matplotlib
Data Engineering · dbt · Airflow

Medallion Pipeline: Airflow + dbt

End-to-end Bronze→Silver→Gold pipeline with 29 automated data quality tests. dbt over DuckDB for modeling, Airflow for orchestration.

Airflow dbt DuckDB Pandas
LLM · RAG · Automation

InforMessi

LLM-powered editorial pipeline that generates and publishes daily Telegram posts about the Argentine national team. Custom anti-repetition RAG system, runs daily via GitHub Actions.

Ollama / Groq RAG Telegram API GitHub Actions
Full-Stack · Next.js · Supabase

SocratIA

Gamified philosophy learning app à la Duolingo. Themed islands, missions with interactive theory and quizzes, XP and daily streaks. Next.js 16 + Supabase + Tailwind v4.

Next.js 16 Supabase Tailwind v4 Framer Motion
Backend · API · CI/CD

FastAPI Payments API

REST payments API with explicit state machine, unit + integration tests, GitHub Actions CI, and auto-deploy to Render.

FastAPI unittest GitHub Actions Render
DL · Finance · NLP

Stock Price Prediction (MSc)

LSTM, GRU and Transformer models for Argentine asset price prediction, combining financial time series with Reddit and news sentiment analysis using BETO.

PyTorch BETO / NLP yfinance Reddit API

02

Teaching & Research

AI instructor, academic researcher, and public speaker at the intersection of cognitive science and machine learning.

// Teaching

AI in Psychology & Neurosciences

Humai · 2023–present

Teaching AI applications for clinical and research settings in psychology and neurosciences.

AI in Mental Health Workshop

Fundación Foro · 2025

Practical workshop on AI tools for mental health professionals.

AI in Psychology

UCCuyo · 2025

AI Tools

Humai · 2024

Social Psychology (Adj. Professor)

Universidad Favaloro · 2024–2025

03

About

I'm a Data Scientist with an unusual background: I hold a degree in Psychology with a specialization in Neurosciences (Universidad Favaloro), and I'm currently completing a Master's in Artificial Intelligence at UdeSA. That combination shapes how I design products and research.

My core interest is the intersection of AI and mental health, using machine learning to make sense of brain signals, clinical data, and behavioral patterns. I've co-authored indexed publications in Neuroscience and Behavioral Science, and I actively teach AI for psychology and clinical settings.

Today I work as Neuroscience and AI Specialist at Humai, leading educational strategy and advanced AI training. I'm also a freelance data analyst at Universidad Favaloro's Vocational Psychology department, which keeps me close to both production systems and applied research.

Technically I live in the space between Data Engineering (pipelines, dbt, Airflow), Applied ML (PyTorch, LLMs, RAG), and product (Next.js, FastAPI). I care about code that ships, and documentation that helps the next person understand why.

Languages
Python SQL Bash JavaScript
AI / ML
PyTorch scikit-learn LLMs / RAG Optuna AI Agents
Data Engineering
Airflow dbt DuckDB Pandas Polars Power BI Looker Studio
Neuroscience Tools
FSL fMRI EEG
Backend / Tooling
FastAPI Next.js GitHub Actions Docker

04

Experience & Education

Experience

AI and Neuroscience Mentor

Humai

2026 — Present

Freelance Data Scientist

Universidad Favaloro

2026 — Present

Academic Director

Humai

2023 — 2026

Quality Control Manager

Further Languages

2020 — 2022

Associate Researcher

Universidad Favaloro

2019 — 2023

Education

Master's in Artificial Intelligence

Universidad de San Andrés (UdeSA)

2025 — Present

Bachelor's in Psychology (Neurosciences)

Universidad Favaloro

2018 — 2024

Data Scientist Program

Coder House

2022 — 2023

High School

Colegio Nac. Bs As

2012 — 2016

Key certifications

AWS Certified AI Practitioner

AWS

2025

LLMs & Deep Learning

Humai

2025

Data Science

Coder House

2023

Big Data

Codo a Codo

2022

05

Get in touch

Open to remote or relocation. Roles in Data Science, AI, and ML Research.

Email
lic.mesch@gmail.com
LinkedIn
/in/sebastian-mesch-henriques
GitHub
@SMESCH1
Location
Buenos Aires, Argentina · open to remote & relocation