Riddhish Bhalodia

AI Researcher

Santa Clara, California, United States10 yrs 9 mos experience

Highly StableAI ML Practitioner

Key Highlights

Expert in multimodal LLM post-training and evaluation.
Proven track record in healthcare AI and medical imaging.
Strong background in AR/VR technologies and 3D generation.

Stackforce AI infers this person is a Healthcare and AR/VR specialist with strong AI and machine learning expertise.

Contact

Skills

Core Skills

Machine LearningComputer VisionMedical ImagingVirtual Reality

Other Skills

AR face trackingmodel developmentevaluation systemsCodec avatars3D generationLLaMA post trainingdata pipelinesagentic workflowsprobabilistic modelingshape analysis2D/3D biomedical image modalitiesdeep learningself-supervised modelssemi-supervised modelslow-resource learning

About

I am a research scientist at Gen AI@ Meta with focus on multimodal LLM post-training. Prior to that I was at Meta Reality Labs. My work is broad spanning all areas of ML models ranging from LLM like Llama as well as 3D computer vision and graphics. Prior to Meta I graduated with my PhD from University of Utah where my thesis focus was on statistical shape modeling for healthcare AI applications.

Experience

10 yrs 9 mos

Total Experience

3 yrs 5 mos

Average Tenure

5 mos

Current Experience

Luma ai

Research Scientist

Jan 2026 – Present · 5 mos · Palo Alto, CA

Nvidia

Research Intern

May 2020 – Aug 2020 · 3 mos

To improve disease localization performance on chest X-Ray images, we utilized the medical reports as an additional data stream. We designed a system that refines the initial bounding box selection using a cross attention-based similarity metric between ROI and text features. The model is trained using contrastive loss based on this cross-modal similarity. Additionally, we also performed characterization of the disease by identifying the keyword attributes, which helps in several applications like severity characterization of pneumonia in Covid-19 patients. Our efforts resulted in this work being published at MICCAI - 2021

disease localizationchest X-Ray imagesmedical reportscross attention-based similarity metriccontrastive losscross-modal similarity+1

University of utah

Research Assistant

Aug 2016 – Aug 2022 · 6 yrs · Salt Lake City, Utah, United States

My research broadly lies in the domain of healthcare AI with a particular emphasis on probabilistic modeling and shape analysis. I work with various 2D/3D biomedical image modalities and utilize machine learning and deep learning approaches to derive clinically relevant and interpretable insights from them. My research utilizes and develops on self- and semi-supervised models, low-resource learning, generative modeling, and representation learning. I extensively collaborate with clinicians and biomedical researchers for my cross-domain projects.

probabilistic modelingshape analysis2D/3D biomedical image modalitiesmachine learningdeep learningself-supervised models+6

Wipro

Research Intern

May 2015 – Jul 2015 · 2 mos · Bangalore, India

During the internship I was tasked with prototyping haptic based functionality in Virtual Market Platform developed at Wipro, CTO. I was deeply Involved in developing VR graphics, and establishing communications between VR interaction using Leap Device and the haptic glove programmed using arduino.

haptic functionalityVR graphicsLeap Devicehaptic gloveArduinoVirtual Reality