Skip to content
View guinik's full-sized avatar
1 follower · 3 following

Block or report guinik

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
guinik/README.md

Hi 👋, I'm Alex Esteban Folch !

ML Engineer & Data Scientist 📍 Barcelona, Spain

LinkedIn LinkedIn • Email [email protected]

🧰 Skills & Technologies

Python C++ PyTorch TensorFlow AWS PostgreSQL Kubernetes Docker Firebase SQL NoSQL CI/CD Computer Vision NLP Data Visualization

📌 Pinned Repositories

The repos listed here are older (some 2+ years old) and mostly simple experiments. But will try to make up some time for small data sciency projects.

🎓 Education

MSc
Universitat Oberta de Catalunya

BSc
Universitat de Barcelona

💻 Projects

Most of my work is private due to project confidentiality, so only a few examples are listed here.

  • Pandaroo — AI Fitness App Personalized workout planning app for people who find the gym intimidating. Built with React Native (TypeScript), uses RAG and LLM pipelines to generate adaptive training plans tailored to the user's fitness level, goals, and preferences.

  • GlobalFunders — Entrepreneur-Investor Matching Platform Platform connecting ambitious founders with investors and funding finders. Built with React and Supabase. Focused on making meaningful early-stage partnerships accessible.

  • Kubernetes-ML-Microservices - ML Inference Platform on Kubernetes Two FastAPI microservices serving a Groq LLM, containerised with Docker and deployed on Kubernetes with HPA, PodDisruptionBudgets, rolling updates, and full AWS EKS infrastructure provisioned via Terraform.

  • BSc Final Project: Restricted Boltzmann Machines for Quantum Many-Body States View PDF Implemented RBMs to model quantum many-body states. Applied Stochastic Reconfiguration, Monte Carlo sampling, and studied Quantum Fisher Matrix spectra and entanglement entropy.

  • Master Thesis: Portfolio Optimization with NLP & Financial Time Series — PRIVATE Applying Reinforcement Learning and NLP on financial news and historical time series to optimise portfolio allocation strategies using deep learning and statistical methods.

🌍 Languages

Spanish
Catalan
English

⚡ Fun Fact

  • I enjoy building AI apps from scratch, exploring new ML architectures, and optimizing code for high-performance systems.
  • Coffee and code fuel my creativity ☕💻

Pinned Loading

  1. FastAPI-Docker-Manager FastAPI-Docker-Manager Public

    A clean FastAPI backend with React frontend for Docker image & container management, built to demonstrate modern architecture patterns and CI/CD.

    Python 2

  2. MNIST-Pure-C-Implementation-with-Autograd MNIST-Pure-C-Implementation-with-Autograd Public

    A C++ implementation of a simple neural network for MNIST with autograd-like support and computational graph topological sorting, translated from a C reference implementation.

    C++ 1

  3. TFG-Restricted-Boltzmann-Machines-and-their-Information-Geometry-in-Quantum-Many-Body-Problem TFG-Restricted-Boltzmann-Machines-and-their-Information-Geometry-in-Quantum-Many-Body-Problem Public

    BSc Final Project (TFG) – Restricted Boltzmann Machines and Information Geometry in Quantum Many-Body Problems. Includes report PDF.

    1

  4. GiveMeSomeCreditDataset GiveMeSomeCreditDataset Public

    End-to-end workflow predicting borrowers likely to default (90+ days past due) within 2 years. Includes data cleaning, EDA with visualizations, model pipelines for Logistic Regression, Random Fores…

    Jupyter Notebook 1

  5. Image-Classification-with-Transformers Image-Classification-with-Transformers Public

    A Transformer-based model for classifying dog breeds, using attention mechanisms to capture subtle features and improve image recognition accuracy.

    Jupyter Notebook 1

  6. Translation-using-Transformers Translation-using-Transformers Public

    Using the Transformer architecture from the Attention Is All You Need paper to build a model that translates text from Spanish to Portuguese with high accuracy and efficiency.

    Jupyter Notebook 1