- Currently building Jobster and developing Project Lifeline (real-time fall detection using FastAPI & MediaPipe)
- Exploring classical machine learning and statistical models
- Interested in model evaluation, metrics, and visual analysis
- Building RESTful backend APIs using FastAPI and Flask
- Hands-on with schema validation & data modeling using Pydantic
- Experience with containerizing applications using Docker
- Regular problem solver on CodeChef, LeetCode, and Codeforces
Pinned Loading
-
CXR-MultiQuant.
CXR-MultiQuant. PublicA Multimodal Deep Learning app for medical triage. Fuses computer vision (DenseNet-121) for Chest X-Ray analysis with NLP (ClinicalBERT) for unstructured clinical notes. Features a React glassmorph…
Jupyter Notebook 1
-
CNN_kernel_response_visualization_template_code
CNN_kernel_response_visualization_template_code PublicVGG16 Feature Map Visualizer: A Computer Vision explainability tool built with Keras and TensorFlow. Inspects the internal layers of a deep CNN by extracting and visualizing raw filter weights and …
Jupyter Notebook
-
VectorForge
VectorForge PublicAn end-to-end Vector Database & RAG pipeline engineered natively in C++. Features a custom HTTP server, algorithmic benchmarking (HNSW, KD-Tree, Brute Force), and multiple distance metrics. Integra…
C++
-
Project_lifeline
Project_lifeline PublicProject Lifeline: A real-time AI fall detection system built for elderly care. Analyzes human posture using computer vision (OpenCV/MediaPipe) to detect emergencies. Features a human-in-the-loop co…
JavaScript
-
insurance_premium_prediction_model
insurance_premium_prediction_model PublicInsurance Premium Category Prediction System: A full-stack ML application predicting premium risk tiers. Features domain-driven feature engineering (BMI, lifestyle, city tier) strictly validated vi…
Jupyter Notebook
-
cuLA
cuLA PublicForked from inclusionAI/cuLA
CUDA kernels for linear attention variants, written in CuTe DSL and CUTLASS C++.
Python
If the problem persists, check the GitHub status page or contact support.


