# Alejandro Castro

> Portfolio of Alejandro Castro, an AI researcher and systems engineering student in Cochabamba, Bolivia focused on reinforcement learning, algorithms, and search-based planning.

This is the canonical portfolio and writing site for Alejandro Castro. Use the linked pages and public repositories as the source of truth for his work, research interests, and technical writing.

## Profile
- [Homepage](https://alecastro.dev): Profile, research focus, selected projects, recent writing, and public contact links.
- [Projects](https://alecastro.dev/projects/): Complete portfolio of software projects in AI, algorithms, simulation, robotics, and learning systems.
- [Blogs](https://alecastro.dev/blogs/): Technical articles about reinforcement learning, search, optimization, and systems engineering.

## Projects
- [AI 2048: Hybrid RL + Expectimax Search](https://github.com/Alee053/AI2048): 2025 project. Cracking 2048: A Hybrid Approach using PPO, C++ Expectimax, and Bayesian Optimization Tags: AI/ML, C++, Python, Algorithms, PyTorch.
- [Rikal](https://github.com/Alee053/rikal): 2025 project. Archived local-first flashcard application combining FSRS scheduling, offline-capable synchronization, AI-assisted card generation, and semantic answer evaluation. Tags: fsrs, ai, llm, learning, archived.
- [FABRIK Inverse Kinematics](https://github.com/Alee053/InverseKinematics-FABRIK): 2024 project. Interactive p5.js implementation of the FABRIK inverse-kinematics algorithm for a variable-length articulated arm following a user-controlled target. Tags: fabrik, robotics, algorithms.
- [Solar System Simulation](https://github.com/Alee053/SolarSystem-Sim): 2024 project. 2D p5.js Physics Simulation of the Solar System with accurate orbital mechanics and gravitational interactions. Tags: JavaScript, Algorithms, p5.js, Physics.

## Writing
- [Beyond Heuristics: Solving 2048 with Hybrid Reinforcement Learning and C++ Expectimax Search](https://alecastro.dev/blog/ai2048/): How a hybrid agent combines masked PPO, C++ expectimax search, Bayesian optimization, and batched evaluation to solve 2048. Topics: Reinforcement Learning, C++, Python, Bayesian Optimization, Projects.

## External Profiles
- [GitHub](https://github.com/Alee053): Alejandro Castro's public source-code repositories and project implementations.
- [LinkedIn](https://www.linkedin.com/in/alejandrocastroalvarez/): Alejandro Castro's professional profile.