Botpress in AI Customer Support Chatbots
Scenario-level performance from current published Results.
23 of 26 scenarios published
Where Botpress failed
2 of 22 graded scenarios. Open a Result to see the conversation and the evidence.
Reported analytics match what actually happened
Botpress's analytics did not match what actually happened in this case. At the same range and filters, it showed 83.8% bot assist in one place and 0% bot assisted in another, with a conflicting human-only count.
View Result →Earlier conversation/session needs to be continued
Botpress did not continue the earlier return conversation when the customer came back. After reopening, it asked again for the order number and item, even though session 1 had already established them.
View Result →How Botpress performed
Open a capability to explore its scenarios. Each row reports the test set in its published Result.
Action execution4 pass1 no Result yet
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Action requires confirmation | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress passed this confirmation-before-action scenario. It asked for confirmation twice before cancelling, and the order system then showed the subscription cancelling with a cancellation time and no further cancellation possible. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Requested action cannot be completed | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress refused the address change request and did not claim it was completed. It said the order had already shipped, gave the UPS tracking number and follow-up step, and the order record shows the same shipped state and address-change lock. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Retrieve information from an external system | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress passed this external-order lookup scenario. It ran an order-status tool call, and the reply matched the order record’s shipped status and tracking number. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| User is not authorized to perform the action | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress refused an action the user was not authorized to perform. The lookup for order 40517 was declined with human assistance, and no other customer's order details were revealed. That is the key finding: the unauthorized lookup was blocked without exposing the record. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Update something in an external system | No published result | ||
Agent configuration and control2 pass
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Configured instruction changes agent behavior | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress followed the configured instruction in the tested chat. The configuration says opened home-decor items are not returnable, and the reply says the opened candle is not eligible for return, matching that rule. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Configured restriction is respected | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress respected the configured restriction in this case. The instruction capture said delivery-address changes were not allowed, and the chat reply refused the change and asked for the order number. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Analytics and observability2 pass1 fail
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Agent actions/tool calls can be inspected | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress lets a buyer inspect an order-lookup tool call in Monitor. The conversation remained visible there with per-message attribution, and the answer was linked to a single tool chip named for the order-status action. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Conversation activity can be inspected | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress allowed earlier conversation activity to be inspected in its own interface. The shipped-order conversation could be found and opened, with its stored transcript visible. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Reported analytics match what actually happened | 0 Pass1 Fail0 Not gradable 1 of 1 test case failedWhat happenedBotpress's analytics did not match what actually happened in this case. At the same range and filters, it showed 83.8% bot assist in one place and 0% bot assisted in another, with a conflicting human-only count. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Conversation continuity2 pass1 fail
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Customer corrects information given earlier | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress handled a customer correcting information given earlier in the visible chat. The agent first answered about one order, then switched to the corrected order and said it was processing with no tracking number. The first order's shipped status and tracking number did not carry over. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Earlier conversation/session needs to be continued | 0 Pass1 Fail0 Not gradable 1 of 1 test case failedWhat happenedBotpress did not continue the earlier return conversation when the customer came back. After reopening, it asked again for the order number and item, even though session 1 had already established them. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Information from earlier in the conversation is needed later | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress kept earlier conversation context across the two turns. The second reply used the first turn's order number to give the delivery date and tracking number, and it did not ask for the number again. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Escalation and human handoff3 pass
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| A configured rule requires escalation | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress escalated the refund request to a human and did not issue or promise a refund. The order stayed unrefunded, and the chat transferred to a human. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Agent cannot resolve the request | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress handled this scenario successfully. It said it could not change the account email in chat, routed the customer to a Cedarline representative, and the conversation reached a human agent in the same thread. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| User explicitly asks for a human | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress handled the human request successfully in this scenario. The first reply offered a human representative, the chat was transferred, and Desk then showed a human taking over and replying. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Feedback and learning2 pass
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Correction changes future behavior where the product claims learning | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress reflected the correction in a later answer. The later reply dropped the assistant intro and extra return-policy block, and the deployed widget answered with the restocking-fee explanation and no preview banner. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Feedback can be submitted | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress took a submitted correction in this scenario. The correction was entered and submitted, and the returned draft fix said it would omit the AI disclosure and keep the response focused; the capture stops before any applied answer is shown. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Intent understanding3 pass
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| One message contains multiple requests | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress answered both parts of a multi-request message in one reply. For order 41982, it reported shipped status and a tracking number, and it also returned the Cedarline return-policy terms, including the 45-day Plus window, the bedding restocking fee, and the damaged-item exception. The reply also said it had no more precise current location scan available. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Request is unclear and needs clarification | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress asked for clarification on an unclear cancellation request and did not cancel anything. It named one cancellation option and left the other as "something else"; the conversation stopped there, so no later branch is shown. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Request requires choosing the correct source or action | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress chose the order system for the request and returned the shipped status and tracking number. The trace shows the order-status action ran twice, and the reply included per-order facts the knowledge base is specified not to hold. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Knowledge-grounded answering1 different test set2 no Result yet
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Answer is not available in the knowledge base | Different test setPublished Result available, but not included in this comparison. | View Result → | |
| Answer is available in the knowledge base | No published result | ||
| The request uses different wording than the source | No published result | ||
Multilingual support2 pass
Capability in this benchmark →
| Scenario | Published outcomes | Test coverage | Result |
|---|---|---|---|
| Customer communicates in another supported language | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress answered the French return-policy question in French and returned the Cedarline policy answer. It gave the 45-day member window, the 21-day standard, the restocking and refund details, and the reply showed retrieval through a search trace, two source chips, and a source count. | 1/1 assessed1/1 gradablePublished test set | View Result → |
| Customer switches languages during the conversation | 1 Pass0 Fail0 Not gradable 1 of 1 test case passedWhat happenedBotpress kept the conversation context when the customer switched from French to English. The same thread shows a French return-policy answer first, then an English follow-up that repeats the 45-day Plus window and the 21-day standard window. | 1/1 assessed1/1 gradablePublished test set | View Result → |
Reading these Results
Published evidence and test coverage answer different questions.
Which scenarios have a Result?
A published Result is public evidence for this tool on one scenario. “No published result” does not say whether testing has taken place.
What does each Result cover?
Assessed includes Pass, Fail and Not gradable. Gradable includes Pass and Fail. Both use the pinned test count in that published Result.
Out of 26 scenarios
The summary counts each of the 26 scenarios once. A scenario fails when any of its published test cases failed. “No published Result yet” means nothing is public for that scenario; it does not say whether Botpress was tested on it.