Organizations seeking to enhance the effectiveness of generative AI should train AI systems using their own data, building upon foundation models. This approach provides a more relevant context and helps address risks like inaccuracies and intellectual property concerns.


  1. An agritech company trains AI models using thousands of annotated images for each crop disease from its app and field teams.
  2. The company relies on its own sourced and vetted datasets to ensure data accuracy and reliability.
  3. Finetuning base models with their own data enables organizations to uphold transparency, control, and responsible AI adoption.
  4. Improved generative AI adoption can be achieved through standard prompt methodologies, cross-platform support, enhanced tooling, and best practices.