Mansi Garg

AI Researcher

Los Angeles, California, United States5 yrs 9 mos experience
AI EnabledAI ML Practitioner

Key Highlights

  • Expert in building scalable AI systems for real-world applications.
  • Proven track record in healthcare AI and multi-agent systems.
  • Strong background in both big tech and startup environments.
Stackforce AI infers this person is a Healthcare AI and Machine Learning specialist with expertise in scalable AI systems.

Contact

Skills

Core Skills

Machine LearningMedical ImagingResearch SkillsPythonMulti-agent SystemsAi AutomationOn-device AiArtificial IntelligenceSoftware DevelopmentAgile MethodologiesProduct Development

Other Skills

.NET Framework3D SlicerAmazon Web Services (AWS)Artificial Intelligence (AI)Azure DevOps ServerAzure SQLBERT (Language Model)Back-End Web DevelopmentC#C++CLI Tool developmentCMakeCascading Style Sheets (CSS)Continuous Integration and Continuous Delivery (CI/CD)Data Science

About

๐„๐ฑ-๐Œ๐ข๐œ๐ซ๐จ๐ฌ๐จ๐Ÿ๐ญ & ๐๐ฎ๐š๐ฅ๐œ๐จ๐ฆ๐ฆ | ๐Œ๐’ ๐‚๐’ @ ๐”๐’๐‚ | ๐๐ฎ๐ข๐ฅ๐๐ข๐ง๐  ๐€๐ˆ ๐ญ๐ก๐š๐ญ ๐ฌ๐œ๐š๐ฅ๐ž๐ฌ ๐Ÿ๐ซ๐จ๐ฆ ๐œ๐ฅ๐จ๐ฎ๐ ๐ญ๐จ ๐ž๐๐ ๐ž, ๐ฐ๐ข๐ญ๐ก ๐ซ๐ž๐š๐ฅ-๐ฐ๐จ๐ซ๐ฅ๐ ๐ข๐ฆ๐ฉ๐š๐œ๐ญ I design and build AI systems that donโ€™t just work in research, but scale to millions of users and devices. From shipping enterprise-scale AI features at Microsoft to building on-device AI assistants at Qualcomm, and now advancing healthcare AI research at USC - I thrive at the intersection of AI, systems, and real-world impact. โœจ What excites me most: - Making LLMs & multi-agent systems reliable in messy, real-world codebases so developers spend more time creating, less time debugging. - Bringing AI on-device, ensuring privacy & accessibility for schools, hospitals, and underserved communities - without relying on the cloud. - Driving healthcare AI - radiomics, wearables, generative models - to help clinicians diagnose earlier, personalize treatments, and improve quality of life. ๐Ÿ’ก I believe the future of AI isnโ€™t just about smarter models - itโ€™s about tangible benefits to humanity: reducing barriers to healthcare, empowering learners, protecting privacy, and building tools that augment human potential. ๐Ÿš€ Whether itโ€™s automating a 400GB+ codebase, fine-tuning biomedical LLMs, or prototyping an on-device assistant in 48 hours, I bring both big tech rigor and startup agility, with a track record of shipping results. ๐Ÿ‘‰ Open to opportunities in AI-first companies, big tech, healthcare AI, robotics, and embedded AI, applying expertise in RAG, LLMs, multi-agent systems, and on-device AI to solve ambitious problems that matter.

Experience

Qualcomm

ML/AI Engineer

May 2025 โ€“ Aug 2025 ยท 3 mos ยท San Diego, California, United States ยท On-site

  • Built a multi-agent AI automation framework using Python & Langchain for Snapdragon AI stack, streamlining development in an enterprise-level C++/CMake codebase powering Qualcomm Neural Network & SNPE SDKs across 7+ hardware backends
  • Engineered a self-healing, RAG-powered CLI orchestrator that auto-detects and parses code changes, writes GoogleTests, fixes build errors using LLM-in-the-loop, & performs coverage analysis, cutting a 6โ€“7 hour developer task to 10โ€“20 minutes
  • <HACKATHON>
  • Built on-device AI assistant for Snapdragon X-Elite using OCR, ASR, and LLM reasoning via ONNX Runtime to capture screen/audio context & deliver just-in-time suggestions fully offline, enabling journaling & tutoring while preserving privacy
  • Developed NPU-optimized ONNX pipeline with Flan-T5 Small, FAISS, and SQLite for on-device context retrieval & logging
LangChainRetrieval-Augmented Generation (RAG)Self-healing RAGMulti-agent SystemsPython (Programming Language)ONNX Runtime+7

Radiomics lab

Research Assistant @ Radiomics Lab

Feb 2025 โ€“ Present ยท 1 yr 1 mo ยท Los Angeles, California, United States

  • Currently developing end-to-end ML pipelines on renal carcinoma imaging data (CT/MRI), including automated segmentation with nnU-Net and 3D Slicer.
  • Building vascular analysis workflows (VMTK, scikit-image, Napari) to quantify biomarkers such as tortuosity, branching, and vessel density.
  • Extracting radiomic features and integrating them into predictive AI models for tumor characterization and prognosis.
Deep LearningMachine LearningPyTorchMedical ImagingNumPyPandas+1

Tadhealth

Artificial Intelligence Developer Intern

May 2024 โ€“ Jul 2024 ยท 2 mos ยท Los Angeles, California, United States ยท Hybrid

  • Designed an AI pipeline with OpenAI APIs that automated therapist workflows, cutting client response time by 40% and scaling mental-health resource matching for 8K+ users by integrating NLP-driven issue understanding & resource mapping
  • Implemented a web app for therapy using BERT-based machine learning, React, FastAPI, Flask and PostgreSQL
Artificial Intelligence (AI)Machine LearningOpenAI APIFastAPIBERT (Language Model)PostgreSQL+4

Keck school of medicine of the university of southern california

Research Assistant @ Frank OCD Lab

Mar 2024 โ€“ Present ยท 2 yrs ยท Los Angeles, California, United States ยท Hybrid

  • Design and implement the pipeline for data analysis for research on the data collected through participants enrolled in the study for understanding science and treatment for OCD
  • Optimize and scale research processes by managing the MyDataHelps platform, advancing the research group's capacity to study and address Obsessive-Compulsive Disorder (OCD) by eliminating the need for daily maintenance personnel
Research SkillsPython (Programming Language)FlaskLinuxAmazon Web Services (AWS)React.js+2

Microsoft

3 roles

Software Engineer (I and II)

Jul 2020 โ€“ Dec 2023 ยท 3 yrs 5 mos ยท Hyderabad, Telangana, India

  • Worked for Dynamics 365
C#React.jsTypeScriptRedux.jsAzure SQLKusto Query Language (KQL)+3

Software Engineer Intern

Jan 2020 โ€“ May 2020 ยท 4 mos ยท Hyderabad, Telangana, India

  • Worked for Dynamics 365
TypeScriptC#React NativeProduct DevelopmentRedux.jsSharePoint Server+2

Software Engineer Intern

Jun 2019 โ€“ Jul 2019 ยท 1 mo

  • Worked for AzureDevops
C#.NET FrameworkAzure DevOps ServerMicrosoft Power BITypeScriptSoftware Development

Education

University of Southern California

Master of Science - MS โ€” Computer Science

Jan 2024 โ€“ Dec 2025

Thapar Institute of Engineering & Technology

BE - Bachelor of Engineering โ€” Computer Engineering

Stackforce found 100+ more professionals with Machine Learning & Medical Imaging

Explore similar profiles based on matching skills and experience