AI Industry Intelligence

Key Challenges

An influential Chinese think tank has spent 5 years building the most comprehensive database tracking the development of China’s AI industry. The client has been heavily relying on manual desk research to identify relationship data between companies


We developed an end-to-end solution that overcomes the two long-time barriers of scaling the database:

  • using network intelligence to automatically identify business relationships; 
  • using exhaustive search to expand data coverage from global sources

Our custom developed algorithmic process resulted in 10x increase of data source converge, and 90% decrease of relationship building time

Target Data Collection

Exhaustively search data from global sources and expand data coverage over 10 times

Relationship Extraction

Automatically identify entity relationships with 86% accuracy and 90% increase of efficiency

Database Redesign

Restructure the database to enable easy BI and ensure consistent quality over future updates

Our highly automated process relying on complex OSINT and NLP algorithms scanned over 1,200 media sources in 5 languages

The proprietary NLP algorithm identified 4,000+ direct and indirect relationship in 5 different types (technical, business, legal, HR and educational) and in over 20 categories


The technology allowed us to reveal hidden insights of the target industry straightforwardly, and it is ready for a broader range of applications including industry knowledge mapping, comprehensive competitive analysis and industry trend analysis