Comment on page
How to successfully manage an AI Brain rollout?
What does AI Brain on-boarding looks like for a typical CX team
A well-planned implementation can significantly impact the user experience and adoption of AI Brain across the organization. Here we discuss a detailed rollout plan that has worked well at enterprise CX organizations. Remember you can alter this plan per your needs.
Rolling out AI Brain for your Customer Experience org
- Identifying Data Sources: Make a list of all the data sources you would like to train the AI with. At this time do not worry about the complexity/permissions. Just make a full inventory list.
- Prioritize Data Sources: This is often a trade-off between where the most important information exists and the hurdles to get access to them. Speed is more important than content. Yes although this is counterintuitive, the success of any AI project depends on how soon you can deliver value. If a data source is going to take time to get access to, then push it into the future and start working with data that's easier to get.
- Data Ingestion and Initial Training: Using the Twig data ingestion tool to import all data into Twig and train the AI Brain. This generally takes a few passes and you add new data test and repeat
- Identify Beta Team: Make a list of people across different CX teams like support, customer success, solution delivery, onboarding, and sales engineering that would be part of the initial beta team.
- Baseline Key Performance Indicators: Make a baseline of the time it takes different teams to respond to customers. Often these numbers are not available so you may need to have a few conversations to get this data
- Leadership Preview: Once the AI brain is set up, the leadership team takes a quick pass at the platform. This step shows the leadership that the platform is ready to onboard beta users
- Onboard Beta Team: An onboarding workshop that shows all beta users how to use the AI and then asks each team member to try a question and share what they are seeing helps people get beyond their initial fear of using an AI. Set expectations that this is the early days of training the AI so it makes a fun experience for everyone.
- Weekly Leadership Meeting: Discuss feedback, adding new data sources, and steps taken to address feedback
- Bi-Weekly Beta Team Meeting: Discuss user feedback and showcase any new updates released to address user feedback
- Expand Beta Team: Now that the beta team has had a few weeks to experience the AI and the training has improved AI responses. It is time to expand to all users in the initial teams
- Expand the all: Often the AI brain trained by CX users is so powerful that all users can benefit from the AI. At this time you can launch the AI to all employees