AI/DL/ML Engineer

Passively Looking

AI/ML Development Kota, Rajasthan, India

Profile
Contacts

Summary

Project: Medicines, Patient and Doctor details from Handwritten Prescriptions
Developed custom Yolov7 (Darknet Architecture) and HTR models using RNN (Bi-Directional LSTM), CNN (Conv2D) layers, and CTC Loss.
Pipeline: Django framework hosted on a dedicated instance, with custom models dockerized and deployed to Vertex AI Endpoints.
Achieved accuracies of 88% with Yolov7 and 70% with HTR.

Project: Facade, Door and Window Segmentation
Developed a custom Mask RCNN (detectron2) for Semantic Segmentation with ResNet 101 as the backbone, achieving an accuracy of above 90%.

Project: Food Packaging Regulatory Labels Detections and FSSAI Checks implementation
Developed AutoML Vertex AI models to detect regulatory labels on food packaging and utilized Document AI OCR for text extraction.
Ensured compliance with FSSAI norms using custom Python scripts.
Pipeline: Flask framework on Google Cloud Functions, with AutoML models deployed on Vertex AI Endpoint.
Achieved 80.5% accuracy in regulatory label detection.

Project: Forecasting Medicine consumption in the upcoming years
Data Preprocessing Pipeline: Automated data mapping, cleaning, imputation
Model Development: Prediction of medicinal drugs for the next year using CNN-LSTM model with over 85% accuracy based on multiple factors.

Set up an ML Development/Deployment Server in AWS EC2 and GCP VM.

Expectations

Having worked as an AI developer for the past two years, I am now seeking new opportunities to further enhance my skills and contribute to cutting-edge projects in the field. Throughout my career, I have developed a strong passion for AI and its transformative potential across industries. With a solid foundation in deep learning, machine learning, and natural language processing, I am driven to leverage my expertise in developing innovative AI solutions that drive business growth and positively impact society.