We're excited to work with you to optimize your agent!
Please fill out this intake form with as much detail as you have available, as we will use this to figure out the best course of action for our engagement with you. We'll let you know if we need anything else from you!
Does it feed into a background process or database? Does it feed to other agents? Do human users interface directly with it?
Please provide as much information as relevant: daily/monthly usage volume, time distribution of usage, internal/external usage split, human/agent usage split
For example, is the agent actively being rolled out to a wider audience?
If the agent gathers information, how do you think about the precision-vs-recall tradeoff? How important is it for the agent to be grounded / cite its sources / etc? What is the crucial information that must not be missed, as opposed to nice-to-have information?
If the agent takes actions, what are the actions? What downstream systems/stakeholders do the actions affect? Is there a validation system before actions are executed? Do the agent’s actions affect what it does next (e.g. modifying a database it may then read from)?
If the agent produces content, is the content consumed by humans or other agents? Are there any must-haves in the style/formatting/tone of the content?
For example: Run it on every possible input? Roll out to a wider set of users? Produce multiple options in parallel? Search over a larger space?
