CV

Research Assistant

University at Buffalo, Buffalo

• Led the R&D of SLAM, Planning, and Diagnostics modules for the Excavator Automation project, secured with a $750,000 funding from MOOG.
• Designed a Field of View optimization algorithm based on COLMAP, to improve camera feature registrations using motion decoupled sensors resulting in a 10.91% increase in feature detection scores.
• Designed a PointNet based neural network model for LiDAR segmentation delivering an accuracy of 91% and IoU of 0.51 enabling real time Field of View optimization at 2 fps.
• Eliminated the need for cross-compilation of FAST-LIO localization ROS2 ML library through Dockerization, thus ensuring reusability and seamless deployment on simulators and development boards.
• Developed an optimized ROS2 Diagnostic node in Python that enhanced team-wide debugging capabilities, resulting in reduced debugging effort and a 60% increase in productivity.

Product Engineer

LTIMindtree, Mumbai

• Led a four-member team in the development of a flagship business analytics tool; Canvas Insights, that automated the ticket analysis process, delivering actionable insights and generating custom dashboards for 150 CXOs across 50 accounts.
• Designed and developed a RASA NLU with LLM integration query engine for Canvas Insights tool, which became one of the three core features of the tool, contributing to 70% of customer acquisition.
• Designed a On-Demand time sharing Machine Learning server for the tool, saving the team $17,000 annually.
• Developed a seasonality detection/ trend prediction model for time series data in ITSM tickets which resulted in discovery of 17 new actionable insights that were added to the tool.
• Enhanced application performance through Dockerization, PostgreSQL database optimization, achieving a 200% reduction in response time and a 55% increase in user interaction.
• Coordinated with clients and on boarded teams to develop automation workflows that reduced client-side tool deployment time by 66% (from 3 months to mere 2 weeks).

Outreach Head

Xavier Institute of Engineering Outreach

• Conducted 6 workshops for the Empowerment of Scheduled Tribe women and children.
• Undertook Computer Literacy, English and Personality Development workshops.
• These workshops help over 200 tribal women children in keeping up with socienty and ever growing technology.

Creative Director

TEDxXIE

• Designed the marketing brochure and the promotional posters for the event.
• Orchestrated stage and auditorium decoration in strict adherence to TED guidelines, ensuring a visually engaging and professional event setup.

Masters of Robotics

University at Buffalo, Buffalo, NY, USA

Engineering Science (Robotics)

• Relevant Coursework: Robotics Algorithms, Robotic Control System, Deep Learning, Reinforcement Learning, Computer Vision.

Bachelor's of Engineering

Mumbai University, Mumbai, India

Computer Engineering

• Relevant Coursework: Artificial Intelligence, Data Structures, Operating Systems, System Programming and Compiler Construction.

F1/10th Race Winner

University at Buffalo

Achieved first place in an autonomous F1/10th Car race held at University at Buffalo with 60+ participants.

LTI SpotLight Award

LTIMindtree

Received Spotlight Award for my outstanding contributions to Canvas Insights.

Best Project

Xavier Institute of Engineering

Won Best Project for my AI Pet robot project at the annual project competition held at XIE.

Best Project in Computer Science and IT

Symbiosis Pune

Won Best Project for my AI Pet robot project at the annual project competition held at Symbiosis Pune.

Hardware architecture and implementation of an AI pet robot

International Journal of Applied Sciences and Smart Technologies

This paper focuses on thehardware aspect of the robot and the working of the Core program (Kernel) of an AI pet robot. It further discusses the algorithm that was implemented in this architecture and compares the final version with other variations. Our work was able to createa fully functioning AI pet robot by using just cheap commercially available development boards, motors and a Desktop PC.

OpenCV Image Processing for AI Pet Robot

International Journal of Applied Sciences and Smart Technologies

The functions of the robot discussed in this paper are face tracking, emotion recognition and a colour-based follow routine. Face tracking allows the robot to keep the face of the user constantly in the frame to allow capturing of facial data. Using this data, emotion recognition achieved an accuracy of 66% on the FER-2013 dataset. The colour-based follow routine enables the robot to follow the user as they walk based on the presence of a specific colour.

Stone, Paper, Scissors Mini-Game for AI Pet Robot

International Journal of Applied Sciences and Smart Technologies

This paper discusses the mini-game aspect of the robot. The game has been developed by using VGG16 convolutional network for identification of the action performed by the user. To improve the accuracy we have made use of background subtraction which gives removes all the unwanted objects from the background and gives a simple cutout of the users hand.

Autonomous F1/₁₀th Car

Developed RRT, AMCL, Dead Reackoning and PID Controller on a 1D lidar enabled F_1/10th car to win first place in an autonomous car race.