Active Project

Status: ๐Ÿ”ต Active - In Development

Description: A neural architecture designed for efficient semantic similarity comparison between queries and documents. This project implements a dual encoder model with in-batch negative sampling for information retrieval tasks.

Technologies: PyTorch, Transformers, FAISS, FastAPI

Features:

  • Query and document encoding into shared embedding space
  • Cosine similarity scoring for relevance ranking
  • Scalable vector search with FAISS indexing
  • REST API for deployment

Current Milestones:

  • โœ… Model architecture design
  • โœ… Training pipeline implementation
  • ๐Ÿ”„ Hyperparameter optimization
  • โณ Production deployment

Links:


Finished Project

โœ… Mathematical Foundations for AI Course

Status: โœ… Completed - Delivered

Description: A comprehensive course covering essential mathematical concepts for artificial intelligence, including linear algebra, calculus, probability, and optimization techniques.

Technologies: Jupyter Notebooks, LaTeX, Python, NumPy, Matplotlib

Key Achievements:

  • Created 12 interactive modules with 50+ exercises
  • Developed visualization tools for matrix operations
  • Built practical examples connecting math to ML algorithms
  • Received 100+ student enrollments in first month

Deliverables:

  • ๐Ÿ“š Course textbook with 200+ pages
  • ๐ŸŽฅ Video lectures covering 15+ hours
  • ๐Ÿ’ป Code repository with working examples
  • ๐Ÿ“ Assessment materials and solutions

Links:


Project Highlights

ProjectStatusTechnologiesTimeline
Dual Encoder for Semantic Search๐ŸŸข ActivePyTorch, FAISS2025-2026
Mathematical Foundations for AIโœ… CompletedPython, LaTeX2024-2025