Overview
In the corporate planning office of a broadcast and software solutions company, I designed an integrated Vision AI platform and a data business model from scratch to find the company’s next growth engine.
Problem
The company’s existing product lineup alone had a clear revenue ceiling. It needed a recurring, data-driven revenue model rather than one-off SI projects.
Approach
I first designed the AI training data pipeline (collection, preprocessing, auto-labeling), then layered deepfake detection, C2PA (content authenticity verification), and VPP (virtual product placement) on top, step by step. I defined the data distribution and sales business model by researching companies on both the supply and demand sides of data, and also identified key technology partners in areas such as invisible watermarking.
Results
Delivered the integrated Vision AI platform architecture and service structure diagrams, established a data distribution and sales business model based on the AI training data pipeline, and won executive approval for the business.
Lessons Learned
A new business that only lists technologies doesn’t get approved. Decision-makers act only when you map out, step by step, how data comes in and what value goes out.