You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The goal is to refine this task into smaller sub-tasks which might include a review of existing approaches (with limitations), creating a proposal, implementing the solution.
User story
Users should be able to create a pipeline that can answer user queries by using previous queries, previous user responses, as well as previous outputs from tools. The pipeline needs to figure out the best tool to answer a specific user query, and it needs to be easy to later debug/visualise what tool was used.
It needs to support the following conversation patterns:
App can ask for input from the user, validate it, and has context on current tool.
Tools available: check_balance(account_id), add_payment_dollars(account_id, amount), get_account_details(email)
User: what is my account id?
_App: get_account_id(email)_
App: I need your email.
User: My email is xxx.
_App: get_account_id(xxx) ==> error_
App: Sorry, this is not a valid email, please provide a new one.
User: xxx@gmail.com.
_App: get_account_id(xxx@gmail.com)_
App: Your account is 1234.
App remembers previous user inputs.
Tools available: check_balance(account_id), add_payment_dollars(account_id, amount), get_account_details(email)
User: what is my account id? My email is xxx@gmail.com
_App: get_account_id(xxx@gmail.com)_
App: Your account is 1234.
User: what is my balance?
_App: get_balance(1234)_
App: Your balance is $1000.
The text was updated successfully, but these errors were encountered:
mrm1001
added
P1
High priority, add to the next sprint
topic:agent
P2
Medium priority, add to the next sprint if no P1 available
and removed
P1
High priority, add to the next sprint
labels
May 2, 2024
Goal
The goal is to refine this task into smaller sub-tasks which might include a review of existing approaches (with limitations), creating a proposal, implementing the solution.
User story
Users should be able to create a pipeline that can answer user queries by using previous queries, previous user responses, as well as previous outputs from tools. The pipeline needs to figure out the best tool to answer a specific user query, and it needs to be easy to later debug/visualise what tool was used.
It needs to support the following conversation patterns:
App can ask for input from the user, validate it, and has context on current tool.
App remembers previous user inputs.
The text was updated successfully, but these errors were encountered: