AI Engineering - Raffael

Project’s dates

• Start Date - 25/04/2025
• Estimated End Date - 09/06/2025
• Actual End Date - 05/06/2025

Project Idea Definition

Kaleido Search

AI-Powered Personalized Product Discovery Engine

Allow users to describe their needs and preferences of desired products, and use AI combined with RAG to find semantically similar items from an e-commerce catalog.

Product catalog contains products from different stores (e.g. Amazon, Otto, Ikea). Product descriptions are embedded in vector store, DB stores products with URL, Title, Price, Thumbnail, Store, Metadata, and reference to document in vector store.

Users can enter natural language queries (e.g., "a stylish and comfortable armchair for a small living room”). Optionally, the user can respond to LLM generated clarification questions about the query, to guide the user.

User activity, such as queries, bookmarks and product clicks, is tracked to provide product recommendations. Users can bookmark products, see their activity, and set preferences like LLM model to use or length of product descriptions.

Login via Google.

The LLM can do:

• Search Query Expansion
• Identify & Clarify Needs
• Product Description Enhancement
• Personalized Product Recommendations
• Provide explanations for specific characteristics, f.ex. manufacturing Materials

UI Mock-up

GitHub Repo

Project Repo

GitHub - Raffa139/KaleidoSearch: AI-powered product discovery engine using semantic search.

AI-powered product discovery engine using semantic search. - Raffa139/KaleidoSearch

github.com

Experimenting Repo

GitHub - Raffa139/VectorEmbeddings: Experimenting with vector embeddings, LLM's, and LangChain

Experimenting with vector embeddings, LLM's, and LangChain - Raffa139/VectorEmbeddings

github.com

Tech Stack

Python, FastAPI, LangChain, SQLAlchemy, TypeScript, React, Vite, PostreSQL, ChromaDB, Docker

Database Schema

LangChain Pipeline

Implemented inside the experimenting GitHub Repo

LangGraph Agent

Comparison Table

Comparison Table

Project Video Recording

Project Slides

[Optional] Screenshots

Main Topics of Learning Materials

• Fundamentals of AI, Machine Learning, LLM’s
• Evolution of AI - History of Machine Learning
• What changed with LLM’s - ChatGPT Revolution
• Current market leads and differences between them
• Natural Language Processing
• Data Preprocessing
• Data Cleaning
• API Integration with AI models
• Structured Output
• Prompt Engineering
• GenAI Ethics
• Advanced GenAI
• Retrieval Augmented Generation (RAG)

External Optional Resources

Examples of projects

RagLangBot Project Overview

Examples of bad projects

V1 Milestones

V2 Milestones