1. Created a web app (Angular framework) with flask for capturing and extracting frame temple information from images, enabling users to snap photos with smartphones and utilize OCR for detailed temple data extraction, resulting in a 40% reduction in manual data entry time and a 25% increase in data accuracy.2. Built a microservice using the Ollama framework that allows users to interact with Medllama (LLM specifically tailored for medical question answering (QA) and other healthcare-related tasks) for disease diagnosis by inputting their symptoms, achieving a 30% increase in diagnostic accuracy. Additionally, contributed to a Dockerized solution for the GPT builder within an Azure environment, improving deployment efficiency by 40%.

1. Administered support for a hapi.js backend, integrating MongoDB aggregation pipelines for conditional frame scanning cost calculations, resulting in a 30% increase in query efficiency.2. Successfully implemented Segment Anything Model (SAM) for accurate segmentation of glasses from users’ faces, achieving a 95% accuracy rate. Developed a robust solution that seamlessly adapts to various devices including iPhone, Android phones, and webcams, precisely isolating glasses from facial images captured on different devices.

1. Implemented headpose estimation technology in C++, ensuring precise alignment of the device with the user’s head position.2. Played a key role in integrating the headpose estimation feature, which led to a 20% increase in user satisfaction and a 15% improvement in overall functionality. Actively participated in the coding process, successfully pushed the developed code to production, and created a monitoring dashboard to gain insights and receive alerts on critical bugs and issue.

Worked on recognition of COVID-19 patients with help of chest X-ray images using Deep learning model .

Sponsorship team leader