Victor
Figueroa
Prestrud
Download CVContact Me

AI & Data Science

Victor Figueroa
Prestrud

Building intelligent systems - from reinforcement learning to production web apps.

View My Work
Scroll
Victor Figueroa Prestrud

About Me

Victor Figueroa
Prestrud

Hello! I'm a passionate and results-oriented professional with a strong foundation in data science, artificial intelligence, and natural language processing. My journey in tech has been driven by a curiosity to solve complex problems and build intelligent systems that make an impact. I have experience in developing machine learning models, crafting NLP solutions, and building full-stack web applications. I'm always eager to learn new technologies and take on challenging projects.In my free time, I enjoy exploring new datasets, contributing to open-source projects, and staying updated with the latest advancements in the AI field. Let's connect and build something amazing together!

3+

Years Experience

5+

Projects Built

AI

Specialisation

What I've Built

Projects

Kravbank: AI for Public Procurement (Bachelor Thesis)

01

Kravbank: AI for Public Procurement (Bachelor Thesis)

Developed a Retrieval-Augmented Generation (RAG) system using MongoDB, FastAPI, and LLMs for public procurement tenders. A detailed report on this thesis project is available. Focus included automating requirement extraction and improving tender drafting.
MongoDB
FastAPI
Supabase
Python
RAG
LLMs
HNSW
N8N
Playwright
Deep Learning for Image Classification (PGR207)

02

Deep Learning for Image Classification (PGR207)

Trained and evaluated MLP and CNN models for MNIST and CIFAR-10 image classification, experimenting with data augmentation, optimizers, and learning rate schedulers.
PyTorch
Python
CNN
MLP
Albumentations
Source
Predictive Analytics: Time Series Forecasting (PGR304)

03

Predictive Analytics: Time Series Forecasting (PGR304)

Applied SARIMA models for time series forecasting, focusing on trend/seasonality analysis, model selection, and clear data visualization. A report is available.
Python
SARIMA
Pandas
Statsmodels

04

Advanced AI Architectures (RL & PSO)

A comprehensive implementation of two core AI paradigms: A Q-Learning agent for autonomous warehouse navigation and a Particle Swarm Optimization (PSO) algorithm for complex function minimization.
Warehouse Robot (RL)

Warehouse Robot (RL)

Q-Learning agent navigating a stochastic environment with a 67% success rate.

Reinforcement Learning
Q-Learning
Particle Swarm Optimization

Particle Swarm Optimization

Evolutionary algorithm optimizing complex multimodal functions like Rastrigin.

Swarm Intelligence
Global Optimization
Source

Live Demo

Q-Learning Agent

An interactive demo of Q-Learning. The same algorithm at the core of my RL & PSO project, rebuilt in TypeScript to run live in the browser. Tune α, γ, and ε and watch the dog learn to find the treat.

🦴
🧑
🧑
🧑
🧑
🐕
🐕 Dog🦴 Treat (+10)🧑 Human (−3)Wall

Episode

0

Successes

0

Success Rate

0%

Hyperparameters

Learning Rate (α)0.30

How quickly the agent updates its Q-values

Discount Factor (γ)0.90

How much future rewards are valued vs immediate

Exploration (ε)0.30

Probability of taking a random action (explore vs exploit)

Animation Speedmed

How it works

The dog 🐕 starts at S and must reach the treat 🦴 while avoiding humans 🧑 and walls. Each step it chooses to explore (random action) or exploit (best known action) based on ε.

After each step the Q-value is updated via: Q(s,a) ← Q(s,a) + α·[r + γ·max Q(s') − Q(s,a)]

Toggle Q-Values to see the learned policy. Arrows show the best action per cell, brightness shows how valuable each state is.

What I Work With

Skills

01

Data Science

  • Python
  • Machine Learning
  • Data Analysis
  • SQL & Databases

02

Artificial Intelligence

  • Deep Learning
  • Neural Networks
  • Computer Vision

03

Natural Language Processing

  • Text Classification
  • Sentiment Analysis
  • Speech Recognition

04

Web Development

  • React & Next.js
  • Node.js & Express
  • API Development

Experience & Education

Resume

Curriculum Vitae
Download

PDF preview not available in this browser.

Download PDF

AI Assistant

Ask About My Resume

Let's Talk

Get In Touch

Have a project in mind or want to collaborate? Pick the channel that works best for you.

Email

vctrfgr1@gmail.com

Phone

+47 462 96483

LinkedIn

victorfigueroa001

GitHub

QuasimodoCodes

Or send a message directly