AI Product Leadership for Manufacturing QA Automation

The Challenge

A leading solar panel manufacturer faced QA bottlenecks across three factories producing over 200,000 panels daily. Manual inspection was slow, error-prone, and difficult to scale consistently.

Approach & Execution

Project Overview

A major Chinese solar manufacturer needed to automate and significantly improve its quality assurance process for photovoltaic panels across multiple high-volume factories.

My Role & Work Done

Led the product management function for the AI visual inspection tool initiative.
Conducted on-site analysis to understand factory workflows, operator needs, and technical constraints.
Defined user requirements, prioritizing features based on impact and feasibility.
Authored detailed user stories and acceptance criteria for the AI model and user interface.
Designed the UX flow and created UI mock-ups for the inspection tool, focusing on ease of use for factory operators.
Established benchmark metrics (speed, accuracy) to evaluate the AI system’s performance against manual inspection.
Collaborated closely with the engineering team throughout the development lifecycle.

Deliverables

Comprehensive product requirements document (PRD).
Prioritized backlog of user stories.
UX concepts and detailed UI mock-ups.
Defined performance benchmarks and success criteria.

Solution Overview

Defined the product vision, requirements, and UX for an AI-powered visual inspection system. Translated complex factory needs into actionable user stories and technical specifications for the machine vision model and operator interface. Established clear metrics for accuracy and speed improvements.

Key Results & Impact

Deployed AI solution achieved QA inspection rates >10 times faster than human operators.
Improved defect detection accuracy by 3x compared to the best human inspectors.
Enabled consistent, scalable quality control across high-volume production lines.
Delivered clear product roadmap and specifications guiding successful development and deployment.