Projects
Research pipelines, autonomous-systems work, and full-stack builds, all in one place.

SCAATA
Self-Critiquing Autonomous Algorithmic Trading Agent
An end-to-end reinforcement-learning trading pipeline across 6 equities. LLMs scraped and normalised 79 algorithmic strategies into a 245,000-sample expert dataset, then a RecurrentPPO agent was warm-started via imitation learning.

EcoPlay
Gamified Environmental Learning Platform
Project Admin for a GirlScript Summer of Code 2026 selected platform, featuring mini-games, an Eco Village system, and SQL-backed dashboards.
Volatility-Adaptive Candlesticks
Pattern Recognition under Volatility Regimes
Heikin-Ashi preprocessing adapts to volatility regimes, then a Random Forest classifier evaluates candlestick-pattern predictiveness.

Nomly
Full-Stack Airbnb Clone
A production-style booking platform built solo for a full-stack take-home assignment: real JWT auth with bcrypt-hashed passwords, a booking-overlap algorithm that guarantees zero double-bookings, and a full host CRUD dashboard alongside the guest-facing browse, search and book flow.

Redactly
PII Redaction Tool
A CLI and web tool that finds personally identifiable information in PDFs and text, then writes a redacted document with every PII value swapped for a consistent, realistic fake. Uses a hybrid regex + spaCy NER pipeline, evaluated end-to-end against a real SEBI IPO prospectus rather than synthetic data.
Fake News Detection
NLP Misinformation Classifier
An NLP pipeline that classifies news articles as real or fake, comparing Logistic Regression, SVM, Naive Bayes and Random Forest across accuracy, precision and recall to pick the strongest model.