Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I’m Anerudh Raina, a Computer Vision Engineer with a track record of designing, optimizing, and scaling AI-powered systems across AR/VR, automotive ADAS, and surveillance domains. I enjoy turning complex research into robust, production-ready pipelines, from real-time edge inference to scalable data processing that informs calibration and testing. I’ve built and deployed real-time video analytics using TensorFlow, PyTorch, OpenVINO, and OpenCV, led data ingestion and sensor fusion efforts, and mentored junior engineers. I thrive in interdisciplinary teams and love solving practical problems that bridge research and delivery.…Hi, I’m Anerudh Raina, a Computer Vision Engineer with a track record of designing, optimizing, and scaling AI-powered systems across AR/VR, automotive ADAS, and surveillance domains. I enjoy turning complex research into robust, production-ready pipelines, from real-time edge inference to scalable data processing that informs calibration and testing. I’ve built and deployed real-time video analytics using TensorFlow, PyTorch, OpenVINO, and OpenCV, led data ingestion and sensor fusion efforts, and mentored junior engineers. I thrive in interdisciplinary teams and love solving practical problems that bridge research and delivery.

Anerudh Raina

Data Scientist, AI Engineer, Developer, +2





Hi, I’m Anerudh Raina, a Computer Vision Engineer with a track record of designing, optimizing, and scaling AI-powered systems across AR/VR, automotive ADAS, and surveillance domains. I enjoy turning complex research into robust, production-ready pipelines, from real-time edge inference to scalable data processing that informs calibration and testing. I’ve built and deployed real-time video analytics using TensorFlow, PyTorch, OpenVINO, and OpenCV, led data ingestion and sensor fusion efforts, and mentored junior engineers. I thrive in interdisciplinary teams and love solving practical problems that bridge research and delivery.…Hi, I’m Anerudh Raina, a Computer Vision Engineer with a track record of designing, optimizing, and scaling AI-powered systems across AR/VR, automotive ADAS, and surveillance domains. I enjoy turning complex research into robust, production-ready pipelines, from real-time edge inference to scalable data processing that informs calibration and testing. I’ve built and deployed real-time video analytics using TensorFlow, PyTorch, OpenVINO, and OpenCV, led data ingestion and sensor fusion efforts, and mentored junior engineers. I thrive in interdisciplinary teams and love solving practical problems that bridge research and delivery.

Available to hire

Hi, I’m Anerudh Raina, a Computer Vision Engineer with a track record of designing, optimizing, and scaling AI-powered systems across AR/VR, automotive ADAS, and surveillance domains. I enjoy turning complex research into robust, production-ready pipelines, from real-time edge inference to scalable data processing that informs calibration and testing.

I’ve built and deployed real-time video analytics using TensorFlow, PyTorch, OpenVINO, and OpenCV, led data ingestion and sensor fusion efforts, and mentored junior engineers. I thrive in interdisciplinary teams and love solving practical problems that bridge research and delivery.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Language

English

Fluent

German

Intermediate

Work Experience

Software Engineer E5 - Computer Vision at Meta Platforms Inc. - Deputed by Creospan Inc.

May 1, 2025 - October 26, 2025

Built scalable distributed data pipelines analyzing terabytes of AR/VR telemetry to generate statistical reports on projection accuracy, enabling a 25% improvement in calibration accuracy. Diagnosed and resolved critical race conditions in multi-camera systems (Toshiba Telicam cameras), eliminating persistent failures in projection error measurements across AR devices with Android OS. Programmed and deployed an automated six degrees of freedom (6-DOF) robot arm (Yaskawa Motoman GP7) for the data capture pipeline, enhancing reproducibility and accelerating testing cycles for cross-device validation.

Senior Software Engineer - Computer Vision at Deepsight AI Labs

March 1, 2020 - October 26, 2025

Designed and delivered full-stack software solutions for attendance and safety systems (e.g., firearm and helmet detection), contributing to ~80% revenue growth. Developed and deployed scalable, high-performance YOLOv2-based real-time video analytics pipelines using TensorFlow, Docker, MLflow, and OpenVINO; optimized for constrained hardware and budgets. Engineered robust data ingestion and preprocessing pipelines for CCTV video streams, ensuring consistent data quality for downstream model training and system reliability. Built, tested, and deployed a face recognition-based attendance system on edge devices using dlib, MLflow, Docker, and OpenVINO, delivering low-latency inference and production-grade reliability. Prototyped a cloud-native traffic counting system using AWS Kinesis and Rekognition, showcasing event-driven architecture and real-time data processing capabilities. Mentored junior engineers and interns in software engineering practices and core ML/CV concepts.

Software Engineer - Camera, BU ADAS at Continental Automotive Components Pvt. Ltd. - Deputed by Kritikal Solutions Pvt. Ltd.

April 1, 2018 - October 26, 2025

Improved downstream ADAS algorithms (TSR, Lane Detection) by calibrating fisheye lens distortion using curve fitting. Enhanced image processing pipeline: improved contrast by 15%, reduced SNR by 2% via C++ demosaicing algorithm updates. Developed and productionized a regular polygon detection algorithm, reducing inference time from 21ms to 2ms (~10× speedup), increasing TSR accuracy by 5%. Created a homography-based preprocessing module for pavement sign detection using BEV representation, camera-IMU fusion and SIMD optimization, contributing to ~1% revenue growth.

Data Scientist - Computer Vision at CubeIt.IO

February 1, 2016 - October 26, 2025

Boosted customer engagement by 20% by implementing an automated album recommender system via DBSCAN clustering (scikit-learn in Python) of image timestamps on the Android app Cubeit.