Aditya Ramesh Email & Phone Number

San Jose, Ca, Us

1. Led the on-device Computer Vision initiative for the Smart Edge Video product to bring privacy regulation, compliance, and visual analytic tools. Designed Computer Vision pipelines for automating the management of video data, including the redaction of personal identifiable information and optimization for video feeds.2. Worked on security features for Cisco’s enterprise Gen AI offering for adoption of safe and trustworthy AI. Designing code vulnerability guardrails and customizable guardrails using Small Language Models (SLMs).3. Currently designing Agentic Workflows for browser interactions using Large Action Models.

San Jose, Ca, Us

1. Worked on vision based experiences for eBay’s camera platform. Led the Guided Capture project for automated listing flow using ARKit and 3D object detection. Patent has been filed for the same.2. Worked on Scene Text Detection and Recognition pipeline (OCR service) to aid in fraud and counterfeit product detection including scaling and deployment optimizations.3. Designed a scalable image similarity pipeline using perceptual hashes for counterfeit image identification. Using algorithms like ANNOY and FAISS for large scale vector similarity search.

Santa Monica, California, Us

1. Developed an EfficientNet based hierarchical image classifier with a directed acyclic graph (DAG) hierarchy for labels for fine grained image classification on highly imbalanced data. Improved performance by 0.23 on the F2 metric against the previous model.2. Productionized the classifier using mixed precision training routines to enable model to handle data at scale resulting in 2x decrease in cost and required cluster size.3. Implemented drift detection and uncertainty and representative sampling based active learning pipeline to reduce annotation costs and close the loop to account for model decay in production due to data drift.4. Implemented A/B testing pipeline to calculate inventory loss with bootstrapping.5. Worked on hierarchical fine grained video classifier using Temporal shift module to improve performance on videos with limited training data.

Ann Arbor, Michigan, Us

Worked on understanding human trust in multi-agent AI systems.

Ann Arbor, Michigan, Us

Investigating human driver's takeover readiness in autonomous vehicles using high fidelity UMTRI driving simulator.

Menlo Park, Ca, Us

Ann Arbor, Michigan, Us

Worked on clustering algorithms for large scale autonomous vehicle interaction data.

Bangalore, Karnataka, In

Algorithms for cluster analysis.

Pittsburgh, Pa, Us

Worked on 'high accuracy on-chip diagnosis of Integrated systems'. Using machine learning algorithms like k-nn's and neural networks to perform module level on chip diagnosis.

Worked on intelligent energy conservation algorithms for BLE based indoor tracking system for health care personnel.

Grad Student, Electrical And Computer Science Engineering (Vision / Robotics)

Bachelor Of Engineering (Be), Telecommunications Engineering