Ø Technical Leadership and Major development in Machine Learning based Retail space Project – Automated checkout system that uses computer vision and ml to recognize objects being checked out

Ø .Designed Cascading neural network to solve problem of complexity with this project.

Ø Development on C++

Architected, Designed and co-developed the software part on the router in C++ using STL

o Rest interface layer, Adaptation layer, WIFI adapter

o Communication between layers using message queues, web sockets. MQTT also used.

· Architected and independently developed from scratch in C#/C++/C a multi machine test automation system for automating testing. Also Led team for further downstream implementation.

· Designed, Co-developed the Cloud component that captures WIFI data provided by router.

Anomaly detection using ML(Machine Learning) on AWS Green grass

Architected major framework of project.

Independently made adaptation layer in C++ for openwrt/Linux. Also uses MQTT

interfaced with lower layers on BLE, ModBus , Zigbee

·Architected and Independently developed C++/C# based test automation system for multiple modules.

Cross compiled for Openwrt, Cross Compiled For embedded

ML- Anomaly Detection using Machine learning on AWS Green grass using Neural networks

Cloud based indoor map server hosting multiple indoor maps, with client-side pages based on JavaScript that show indoor map overlaying over google map.

· Developed map server hosting maps made in Java using SOA with rest calls to server which rest on AWS. Used AWS load balancing to scale out system and provide redundancy, reliability

Independently Created module that maps a walkable path in indoor map using Prim’s/Dijkstra’s

Designed, Co-developed the Cloud component that captures WIFI data provided by router.Anomaly detection using ML(Machine Learning) on AWS Green grassIOT project – C++, C, Linux, OpenWrt, Zigbee, Wi-Fi, BLE, C#., cloud component using Java 8, RestIndependently made adaptation layer in C++ for openwrt/Linux. Also uses MQTTArchitected and Independently developed C++/C# based test automation system for multiple modules.Cross compiled for Openwrt, Cross Compiled For embeddedML- Anomaly Detection using Machine learning on AWS Green grass using Neural networks

Cloud based indoor map server hosting multiple indoor maps, with client-side pages based on JavaScript thatshow indoor map overlaying over google map.Developed map server hosting maps made in Java using SOA with rest calls to server which rest on AWS.Used AWS load balancing to scale out system and provide redundancy, reliabilityIndependently Created module that maps a walkable path in indoor map using Prims/Dijkstras

Server used to calculate location based on combination of relative intensity of WIFI signals, cellular locationMade in C++ on Linux, project uses STL data structures extensively – maps, ArrayLists etc.Developed module providing altitude in position using barometry along with current weather data in C++