Business Challenges
A public city transport company transports hundreds of thousands of people per day via its trams and buses. The company wanted to learn about and experiment with using generative AI. Which uses could be of added value to its daily operations and people? Emixa supported the company in taking those first steps and using generative AI.
Keys to Success
Our Data & Analytics consultants know the ins and outs of generative AI. Taking the labour agreements and some of the business IT infrastructure the company already works with, our team developed ChatCAO. This Language model implementation uses Microsoft Azure AI services, Python and Mendix.
One of the challenges in working on Generative AI is to prevent it from 'hallucinating', meaning that it starts making up new information. With the right setup, the ChatCAO language model stays true to the information in the labour agreement documents. Using this setup and showcasing the PDF document along with the corresponding page in which the language model found the information, makes for a reliable solution with trustworthy information.
Results
The company’s employees now have a tool that precisely answers questions about their labour agreements, replacing time-consuming document searches and ensuring consistent information distribution.
Taking the first steps in Generative AI
Happy to help this company take its first steps with Generative AI. It's good to build up these skills in any organisation.
Han Raaijmakers
Emixa