• Boosted employee productivity by 34% by enabling efficient access to information and intelligent responses to user inquiries using LLM-powered chatbot with document retrieval using Langchain and Pinecone • Produced impactful KPI dashboards using Power BI, aiding 65% improvement in data driven decision-making • Implemented CI/CD pipelines using Jenkins to automate the deployment process, ensuring rapid and reliable delivery of software updates.

• Translated business requirements into effective data-driven solutions, leading to 18% optimization in farming strategies • Designed the LGS Grower Subsystem database, laying the foundation for efficient data management • Implemented optimized SQL queries in the Agile Scrum-based development process, resulting in a significant 30% reduction in data retrieval time

• Managed IT support services, troubleshooting, and system operations, successfully resolving over 100 user issues.• Designed and created an internal dashboard to improve the efficiency of employees and streamline operations

• Trained and optimized explainable AI models for satellite and drone monitoring systems using PyTorch, resulting in 30% decrease in model prediction time • Created object detection models for drones for object tracking and surveillance using Yolov7 • Enhanced model accuracy by 20% by contributing to creating, labeling, and manipulating extensive image datasets using Python. Leveraged web scraping for sourcing and OpenCV for manipulation

• Led multiple client projects in NLP and Computer Vision, leveraging PySpark for efficient model training. Applied transfer learning techniques on GPT, YOLO, and BERT using PyTorch to develop tailored solutions • Collaborated cross-functionally to deliver innovative AI/ML solutions, communicating effectively with stakeholders to understand and address client needs efficiently • Improved API response time by 34% by creating ONNX model deployment framework on NodeJS and AWS, streamlining the entire model training process, ensuring scalability, and API generation for production deployment • Reduced deployment overhead by 28% and streamlined development processes by implementing multi-model endpoints on AWS SageMaker using TorchServe, while conducting A/B testing to evaluate deployment effectiveness.

• Drove a significant 36% revenue increase by leading the creation and deployment of website powered by AngularJS and NodeJS • Enhanced user experience and security by integrating advanced components, such as chatbot, fraud detection, and recommendation engine

• Created text recognition and face recognition models and deployed them as APIs. The models were used for internal usage to boost staff productivity.