ML evaluation
Detecting wheat rust from satellite time series, with an evaluation built to catch shortcuts and leakage.
/PROJECTS
Projects
Machine learning and full-stack work, and what each one is built on.
-
Featured
Wheat Rust Detection
A research pipeline and transformer model that detects wheat rust from Sentinel-1 radar and Sentinel-2 optical time series, with a leakage-controlled evaluation against a re-implemented benchmark.
-
Featured
Magnite LLM
A desktop app for downloading, running, fine-tuning and querying language models on your own machine, with a C++/Qt 6 interface and a local Python backend.
-
Featured
AirportOpsEnv
An OpenEnv-compliant RL environment where an AI agent plays an airport ground-operations controller.
-
Brightloop
A subscription management web app where staff manage products, plans and subscriptions, and a billing engine turns active subscriptions into invoices.
-
AI-Powered E-Consultation Suite
A Streamlit app that scores the sentiment of public comments, pulls out key themes, and summarizes long text, aimed at government e-consultations.
-
FinWise AI
A retirement-planning web app for Indian users, combining a Botpress chat agent, a retirement calculator and a monthly email reminder server.
/FOCUS
Where this applies
The kinds of problem the work is about.
/ABOUT
About
Engineering student in Mumbai working on machine learning and software.
- Based in
- Mumbai, India
- Education
- B.E., Thakur College of Engineering and Technology, Mumbai. 2023 to present. CGPA 8.69.