Atin Mathur
Skills & Expertise
Product leadership and technical execution capabilities honed over 10+ years
Product Management
AI/ML Products
Enterprise Execution
Technical Execution
Product Design
PM Toolkit
Featured Projects
Innovative solutions that have made a real impact
Platform.AI
Built a custom image and video annotation platform for GE Appliances, enabling large-scale labeling of food type and doneness across microwave camera feeds. Developed and iterated on food classification and doneness estimation models, enabling real-time smart cooking decisions and improving ML development velocity by ~40% through optimized data and labeling pipelines.
Product Approach:
- • Built custom ML pipelines for specific client needs
- • Focused on enabling real-time smart cooking decisions
- • Optimized data labeling and annotation workflows
Impact:
- • Improved ML development velocity by ~40%
- • Enabled large-scale food classification for enterprise client (GE Appliances)
- • Delivered real-time doneness estimation for microwave systems
Technical Foundation: React, Python, FastAPI, TensorFlow, Google Cloud
Burn.AI
Led early-stage product discovery with a US-based top plastic surgeon for automated burn severity classification. Built a mobile ML-powered PoC to validate technical feasibility and clinical workflows, using findings to guide product viability and investment decisions.
Product Approach:
- • Early-stage product discovery with clinical stakeholder
- • Built mobile ML-powered proof of concept
- • Validated technical feasibility and clinical workflows
- • Guided product viability and investment decisions
Impact:
- • Successfully validated technical feasibility for burn severity classification
- • Mobile PoC demonstrated real-world clinical workflow integration
- • Findings informed product strategy and investment decisions
Technical Foundation: React Native, Python, TensorFlow, Flask, Google Cloud
ImplantID
Led end-to-end product ownership for an AI-assisted spinal implant identification platform used by Fortune 500 medical device clients (Medtronic), reducing manual identification effort by 4,000+ hours annually and delivering $2M+ in recurring enterprise value. Translated complex clinical, regulatory, and engineering requirements into production-ready mobile and backend specifications, enabling compliant deployment across multi-device workflows.
Product Approach:
- • Led end-to-end product ownership from concept to enterprise deployment
- • Translated complex clinical, regulatory, and engineering requirements
- • Drove roadmap prioritization and cross-functional execution (AI, mobile, backend, clinical stakeholders)
- • Enabled compliant deployment across multi-device workflows
Impact:
- • $2M+ in recurring enterprise value
- • 4,000+ hours of manual effort eliminated annually
- • Deployed to Fortune 500 medical device client (Medtronic)
- • Maintained medical-grade accuracy and reliability thresholds
Technical Foundation: React, React Native, Python, FastAPI, Google Cloud
SpineIntel
Early-stage proof of concept focused on spino-pelvic parameter analysis, developed in collaboration with a top spinal/neurosurgeon. Building ML-powered tools to assist clinical assessment and surgical planning through automated measurement and analysis of spino-pelvic parameters from medical imaging.
Product Approach:
- • Clinical collaboration with top spinal/neurosurgeon to validate requirements
- • Building ML-powered tools for automated spino-pelvic parameter measurement
- • Focused on improving clinical assessment and surgical planning workflows
- • Early-stage proof of concept to validate technical feasibility
Impact:
- • Validating automated measurement approach for spino-pelvic parameters
- • Proof of concept in active development with clinical guidance
- • Exploring ML techniques for clinical decision support in spinal surgery
Technical Foundation: React Native, FastAPI, OpenAI, Langchain, PyTorch, Supabase