Automating electronic board review with AI
Together with TPV, we built an AI-powered platform that automates the technical analysis of electronic boards, cutting a process that once took hours down to just minutes.
Case background
TPV is one of the world’s largest manufacturers of monitors and televisions. With its Brazil operations based in Manaus, the company is globally recognized as a leader in monitors and a major player in the TV market.
At its industrial hub, TPV operates two facilities: one focused on manufacturing SMT electronic boards, and another dedicated to final product assembly.
But before any board reaches the production line, there is a critical and often unseen stage: engineering.
How a board is produced
It was a highly technical, repetitive process that depended heavily on specialist expertise — and in total, it could take more than 18 hours per cycle.
Our mission
To eliminate this bottleneck, we needed a solution built around four core goals:
Compare Gerbers
Automate the analysis of Gerber files across product revisions, removing the need for manual visual review.
Validate BOMs
Automatically identify changes in bills of materials between versions with precision and full traceability.
Generate Feeder Lists
Automate the creation of Feeder Lists used to program SMT machines, replacing a process that previously took up to eight hours.
Smart Reporting
Generate clear, actionable reports so the engineering team can make decisions with confidence.
The solution
We developed a platform powered by two integrated AI agents that automates the entire technical review workflow for electronic boards - from file upload to the final report.
File reading and interpretation
Engineers upload the input files, and the platform automatically identifies and extracts all relevant elements from each one.
Intelligent analysis and comparison
The AI agent correlates elements across product versions, identifying what was added, removed, or changed - including the visual position of each component.
Report generation
A second agent consolidates all identified changes and generates an explanatory report with recommendations for the engineering team.
Results
End-to-end traceability across the entire process.
Fewer errors, less rework, and more consistent results.
Less dependence on individual expertise.