A general purpose recommender metrics library for fair evaluation.

Korean-optimized RAG evaluation toolkit with Kiwi tokenizer, ROUGE metrics, and IR evaluation for retrieval systems (Hit@K, NDCG@K, MRR, etc.)

GMD is a lightweight SaaS metrics dashboard that gives founders instant clarity on their business performance. Track MRR, ARR, churn, CAC, LTV, and more through automated data ingestion, Stripe integration, and CSV uploads in an all in one clean, easy-to-use interface.

Fast implementation of the MRR ranking metric

Tiny, reproducible evaluation harness for RAG systems (recall@k, MRR)

Deployed RAG copilot with RBAC, citations, hybrid retrieval, reranking, and evaluation harness.

This project develops a domain-specific embedding model to enhance document retrieval in AI-powered search systems. It incorporates techniques like synthetic data generation, model fine-tuning, and vector search using FAISS, evaluated with MRR@5 for performance.

Small demo project for evaluating RAG retrieval quality using a simple BM25-style retriever and standard metrics (Hit@k, Recall@k, MRR).

Design, Implementation and Evaluation of various Search Engines

Modeling Monthly Recurring Revenue (ARR) and subscription lifecycles using dbt and SQL. Handles churn, reactivations, upgrades, and downgrades via dimensional modeling.

📊 Track key SaaS metrics like MRR, ARR, and churn effortlessly with this intuitive dashboard for clear business insights.

End-to-end recommendation system that suggests next learning items (courses, exercises, resources) based on student interaction data.Includes preprocessing, baseline + GRU4Rec models, evaluation and a Streamlit demo app.