It is an application that estimates how much a new customer can earn on average and in which segment by applying EDA (Exploratory Data Analysis), segmentation and forecasting methods on persona.csv (customer dataset) using the Streamlit library.

LLM-powered academic coach rule-based grade classification + GPT-3.5 Turbo generates personalized study plans, recommendations, and warnings via a 3-step React + FastAPI flow.

End-to-end analysis of 550k+ European data job postings (2023-2025) using Python, SQL, BigQuery and Looker Studio.

This research evaluates 100 ChatGPT-generated passwords against 10 real-world website policies (e.g., banking, healthcare). Using rule-based validation and metrics (zxcvbn/PasswordStats), it analyzes only "fair" and "strong" passwords

Classify SAP BOM components as plastic / polymer / rubber / non-plastic — rules + ML, with confidence, evidence, and a review queue for ambiguous or contradictory rows.