NxtUp : Recommendations System Design Architecture [Django + SQL + JavaScript]

Demonstrated the System Design Architecture of NxtUp's recommendation Engine. It uses a dataset from hugging face with approx 10k questions. Those questions are converted to Vector Embeddings and queried upon for efficient question picking.

Tech Stack Used :
Frontend : HTML, CSS, JavaScript
Backend : Python (Django)
Database : SQL
AI : ChromaDB (Vector Embeddings & Vector DB) and RAGs
APIs : Ollama 9B Instant (GroqCloud Hosted), GEMMA-4 31B IT (Nvidia Hosted), Codeforces API (platform hosted)
Deployment : Yet to be made
Task Scheduling : Yet to be made

LINKS :
(Paste URLs in browser for now)

📄Research Paper on "Lost in the Middle" : cs.stanford.edu/~nfliu/papers/lost-in-the-middle.tacl2023.pdf
🔗 GitHub: https://github.com/amit-kumbhare/NxtUp
📌 Next: Actual DEMO of NxtUp's Recommendation Engine

Видео NxtUp : Recommendations System Design Architecture [Django + SQL + JavaScript] канала Amit Kumbhare