Hey, I'm Hristo đź‘‹

🤔 Who am I?

Python Chatbot Developer with a background in Automotive Mechanical Design Engineering.

What I’m good at
Working independently, communicating clearly, asking good questions, and solving technical problems systematically.

What I’m like
I enjoy working in a team, understanding in depth how things work, and sharing what I've learnt.

🛠️ Projects

A quick look at some of my work, grouped by how long it took and how much effort it needed.

Multi-month

Project Work for AI Program at SoftUni

  • Jupyter book presenting my three final projects:
    • predicting sumo match outcomes with classical machine learning
    • predictive maintenance for Portugese metro trains
    • exploring the math behind Principal Component Analysis (PCA)
    You'll also find my notes here—both typed and handwritten—from 2022 to 2023.
  • classical machine learning, model training, data science, EDA, data preprocessing, statistics, PCA, predictive maintenance.

Digitize Choir Scores

  • Digitized manual choir scores, generated pdf and mp3 files, and built a static site to host them on AWS using a custom domain. The project includes over 1,000 files, totaling more than 1GB after build (lots of mp3) and around 28K lines of code. Started in June 2024 and still a work in progress.
  • LilyPond to generate sheetmusic and audio, Bash and Python scripting to automate process, JupyterBook frontend and Jinja templating, S3 for storage, CloudFront for HTTPS, CloudFront Functions for basic auth, Route53 for a custom domain, CloudFormation IaaC.

A Week or Two

Chat-with-PDF RAG App

  • Chat-with-PDF app with RAG. Uses local ollama LLM (April 2025).
    • showcase
  • chainlid for GUI, langchain for LLM integration and PDF processing, ollama for LLM server, chromadb for vector database.

Manifacturing Processes NLP

  • Build a semantic search and visualization tool for manufacturing processes. Embed Wikipedia-scraped summaries. Then: similarity search, dimensionality reduction, annotation, deployment as a Streamlit app (July 2023).
  • BERTopic, UMAP, SentenceTransformers, Pigeon, SetFit, Streamlit, Bulk .

Few Days or Less

  • Spring Boot Demo (2025-07): A basic Java Spring Boot CRUD app to showcase Spring Boot, Java, and PostgreSQL.
  • Bulls and Cows (2025-06): CLI game using Python curses
  • Terraform Local Provider (2025-04): Terraform can create files
  • Applied Machine Learning: Foundations (2025-03): Re-wrote and extended the ML Foundations course notebook; course by Matt Harisson in LinkedIn Learning. Showcases: sklearn, polars, method chaining, pipelines, CatBoost, random forest, residual plot, validation curve