Raghav Gupta

Co-Founder

Bengaluru, Karnataka, India9 yrs 3 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Pioneered assistive technologies for differently abled.
Developed real-time sports analytics solutions.
Expert in Computer Vision and Embedded Systems.

Stackforce AI infers this person is a highly skilled AI and Computer Vision expert with a focus on real-time applications.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Machine LearningComputer VisionEmbedded SystemsProgramming

Other Skills

Document Understanding PlatformSmart Animal Assistance and Monitoring SystemVideo Content GenerationAI for Precision Agriculture3D Scene ReconstructionInvoice digitisationData-sheet digitisationProduct catalog enrichmentSoil reports digitisationAutomatic document validationMagazine UnderstandingVizard Platformsynthetic document generation frameworkDocument UnderstandingAI

About

Over the course of my undergraduate studies and professional career I have pioneered assistive technologies for differently abled and developed solutions for real-time Sports Analytics. I have gained experience in the fields of Embedded Systems and Computer Vision. I have come to realisation that there is plenty of untapped data around us and CV/ML is just one such platform that can enable us in using it the right way. I am interested in developing next generation solutions for real world applications in these fields.

Experience

9 yrs 3 mos

Total Experience

3 yrs

Average Tenure

Current Experience

Rally vision

Co-Founder, CTO

Jun 2024 – Present · 2 yrs · Bengaluru, Karnataka, India

Jio

2 roles

Technical Lead

Promoted

Aug 2019 – May 2024 · 4 yrs 9 mos

Architected (i) Document Understanding Platform (ii) Smart Animal Assistance and Monitoring System for Vantara Animal Kingdom (iii) Video Content Generation using synthetic Avatars (iv) AI for Precision Agriculture (iv) 3D Scene Reconstruction for high-quality maps/assets
Managed end-to-end product development/deployment for multiple Reliance business: Invoice digitisation for Retail/Hydrocarbon; Data-sheet digitisation for Refinery; Product catalog enrichment for JioMart; Soil reports digitisation for JioKisaan; Automatic document validation for candidate onboarding for HR; Magazine Understanding for JioMags
Worked closely with the UI team design/develop Vizard Platform, a one-stop platform to cater to semi-supervised document labelling, model demonstrations, job scheduling and performance visualisations.
Designed synthetic document generation framework to create richly labelled document data at scale for all possible model building needs for document understanding. It has supplied >1 million documents for model training
Responsible for (i) Building team-wide processes for development, model management, documentation, planning (ii) Client pitches/management (iii) Designing product road maps (iv) Infra procurement for dev/staging/prod environments (iv) Team building/growth (v) Product documentations

Document Understanding PlatformSmart Animal Assistance and Monitoring SystemVideo Content GenerationAI for Precision Agriculture3D Scene ReconstructionInvoice digitisation+9

Data Scientist, CoE AI/ML

Sep 2018 – Jul 2019 · 10 mos

# Real-time face recognition based attendance pipeline
# Network quantisation using TensorRT
# Spoof detection for face images

Real-time face recognitionNetwork quantisationSpoof detectionComputer VisionMachine Learning

Gamelore inc.

Senior Sensor Engineer/Founding Engineer

Oct 2016 – Mar 2018 · 1 yr 5 mos · Bengaluru, Karnataka, India

We invented a device which performs real-time analysis of tennis game to calculate relevant statistics and give feedback to the players and coaches to improve their game. Our team was able to successfully build a working prototype of the device from scratch.
I was responsible for complete design and development of:
Tennis Court Detection Module and Line Call System
Object Detection Module for contextual understanding of the scene and detection of very small objects (racket, ball, player etc) in outdoor lighting conditions
Tennis Shot Classification and Posture Correction System based on player's biomechanics obtained from 2D skeleton
Camera calibration and triangulation using non-linear optimization
Optimization of all the algorithms for real-time functioning on Jetson TX1/TX2

Tennis Court Detection ModuleObject Detection ModuleTennis Shot ClassificationCamera calibrationAlgorithm optimizationComputer Vision+1

Samsung r&d institute india

Senior Software Engineer

Jun 2014 – Sep 2016 · 2 yrs 3 mos · Bengaluru, Karnataka, India

Functional verification and performance validation for Imaging and Vision Accelerator (IVA) which contained optimized and configurable hardware engines to achieve low latency vision applications.

Functional verificationPerformance validationImaging and Vision AcceleratorEmbedded Systems