|
Basic Assignments
|
Options & Settings
|
Main Time Information
|
||||||||||||||||||||||||||||
|
|
|
|
|
|
Sorry, no photos available for this element of time.
|
|
|
|
Notes:
|
|
The Nxtlinq guys are adding in more logs so that we can track some other things. They are working on a number of things including improved memory and global context stuff. Memory in AI is making training and knowledge stick. We also talked about a number of other topics. Here are a few of them: - The costs for the infostructure and keeping the AI agent up and running. - Talking about global information storage - we really need this, from our team to the bot. - Talking about influencing things upstream and downstream. We need the AI bot to remember and learn, but we also have to be careful about where it gets its information. - Light talks about an audio issue between multiple tabs. The AI agent is still listening in a different tab and responding. The simple work around is the mute button on the microphone. We may look deeper into this, if it becomes an issue. - After the main meeting, Steve, Thomas, and I stayed on the meeting to talk about pricing and future plans. - Being able to train the agent so that it can help with support, setup, and deployment. Figuring out the pain points and helping us as a company as well as our clients and users. - Thomas was telling us some stories about introducing agents for other companies. - Steve was talking about being able to help get the agent up and going and helping with oversight. - Lots of talk about ingestion rates and values (how much we feed it). - Human in the loop to help with the training. - Thomas talking about manual crank windows (roll up/down the windows by hand) vs electric windows (power windows). Ideally, we show our clients and users how awesome it is with AI help. Then, hopefully, they won't want to go back to older tech. - Getting into the BI level (business intelligence level) later on, as we go. - If we train this agent up, it will really help our users and clients with training, consulting, etc. - Here are our steps and phases for working with the AI agent - 1. Navigation and prompts, 2. Setup and training, and then 3. Clear out to consulting. - Small phone call with Steve afterwards to go over our internal plan a little bit. Good meeting. |