Building Intelligence, thoughtfully.

LabBench is a lightweight, browser-based IDE and Progressive Web App for students. It combines a Monaco-powered editor, file explorer, interactive terminal, web preview and AI debugging hints in one workspace, so practical lab work can be done on budget phones or locked-down lab PCs. It was built around real lab-session friction: no personal laptop, restricted software installs and tedious output submission.

• Editor, file explorer, terminal and live preview in one workspace
• AI hints explain why code fails rather than giving the fix
• Installable as a standalone app on Android, iOS, Windows and macOS

• TypeScript
• Monaco Editor
• Supabase
• PWA

Source is public for review; the software is proprietary, all rights reserved.

BloodSignal helps families and care teams organise a blood-bank search by blood group, component, quantity and location. Facility directory records and inventory evidence are kept as separate facts, so a listing is never mistaken for stock. Unknown, stale, unavailable and demo evidence are reported distinctly, results are ranked by evidence quality and freshness before distance, and every result points the user to call the facility to confirm. It was originally developed for the Ziro Hackathon by Susanta Banik and Shreya Bhattacharjee as Team Pragmatrix.

• Missing inventory is shown as unknown, never as zero stock
• Typed requests, optional voice input and prescription-assisted fields, reviewed before searching
• Call-to-confirm and map links; distance is labelled straight-line

• React
• TypeScript
• TanStack Start
• Python
• FastAPI

Team Pragmatrix · Ziro Hackathon.

CodeWithFun Dev Hub is an online bookstore for developers and students offering textbooks, technical books and study material at affordable prices. It was designed, built and maintained by Susanta Banik alongside the CodeWithFun YouTube channel.

A binary classification model built with TensorFlow and Keras that identifies whether a transaction is normal or fraudulent, demonstrating how to build and train a deep neural network for the task.

• Python
• TensorFlow
• Keras

Emotion Detector is a computer-vision tool built with OpenCV, the DeepFace library and TensorFlow that estimates a person's emotion from a face. Other attributes it reports, such as age, are estimates and may not be accurate.

• Python
• OpenCV
• DeepFace
• TensorFlow