Developing and managing scalable MLOps pipelines for Observance, an AI-driven construction management platform.
Creating and integrating advanced automation tools along with next-generation computer vision and sensor fusion algorithms.
Computer Vision Intern - Inkers Technology
#Computer Vision
#Real-time Processing
Orchestrated and built a flexible and robust " Automated Processing Pipeline Management" system allowing for multiple parallel deployments, real-time monitoring of each deployment, and automated services capable of redeploying the pipeline in case of unexpected processing instance terminations.
Reduced cloud resources cost by a whopping 40%.
Redesigned and unified the processing pipeline code-base, providing seamless integration support for various Simultaneous Localization and Mapping (SLAM) algorithms such as Fast-LIO, PV-LIO, R3live, and PointLio and support for all in-house built sensor fusion devices.
Learned to integrate data from diverse RGB cameras and thermal devices into LOAM algorithms, resulting in the generation of information-rich colored point clouds.
Enhanced the defects-objects detection pipeline by implementing an odometry-based frame filtering approach, significantly reducing processing times while maintaining consistent and quantifiable outputs.
Deep Learning Intern - Bosch
#Deep Learning
#Autonomous Vehicles
Worked as a full-time research intern to develop cutting-edge deep learning solutions.
Primary areas of work included designing optimised Computer Vision models object detection and analysis.
Worked on building efficient methods of interpreting Point-Cloud data and LiDAR-based 3D object detection for autonomous driving using Single-stride Sparse Transformers (SST).
Also, experimented on Point-Cloud captioning and LiDAR-to-Image generation, using LidarCLIP.
Experimented on Multi-Model architectures like Contrastive Learning models(SimCLR) combined with Unsupervised learning methods for precised object detection.
Intern - PowerApps & AI - Tech Mahindra
#App Development
#Microsoft PowerApps
Worked as a full-time intern to design an automated framework for an existing application.
Designed an end-to-end AI-based entry system which increased the user interaction time in the app by 3x.
Used company standard methods to perform strategy assessments and implementations.
Participated in project meetings, with technical staff members and analysts.
Community @ AIHub.io
#CommunityPortal
#Knowledge Sharing
Founded AIHub.io, a startup focused on delivering AI solutions to healthcare providers.
Managed product development and successfully secured three major clients within the first year.