Architected and built a scalable Portfolio Management System that revolutionized electric bike technology by extracting, preprocessing, and storing data from multiple sources in Azure Synapse and Azure Dataflow, enabling real-time, data-driven insights.Streamlined data pipelines and improved data quality, boosting availability by 90% and accuracy by 80%.Developed high-impact AI solutions:Document classification system achieving 95% accuracy, streamlining document processing and reducing manual effort by 30% for a large law enforcement agency.Designed and implemented a Kafka-based message queue system, replacing REST-based ML services and improving system scalability by 40%.Architected and built a scalable ML platform with Kafka, model management backend, and user-friendly frontend, enabling rapid model integration and deployment.

Image search system finding similar images with 95% accuracy, enhancing user recommendations and increasing engagement by 20% for a large media and entertainment organization.

Worked on building a YOLOv4 based object detection model for detection of aerial object using live camera based feed Designed and developed Face Recognition based application authentication systemBuilt Face Recognition Training architecture with various customizedparameter for training i.e. detectors, networks and object functionDesigned and developed predictive model with high accuracy.Managing the team of more than 10 people.Involved in technical recruitment process

Designed and developed TMCM Motor Driver API for communication using serial(RS 232,Rs485) and Ethernet protocol.Designed and developed Automation API for various Component with structured data set.R&D on designing and development of automation hardware & software according to project requirement.Interfacing and establishing communication between different hardware using RS232,Ethernet and SPI protocol.