Verified answers
Ground responses in the knowledge scope approved for the pilot.
SupportBot uses approved company knowledge to prepare grounded answers, show the evidence behind them, and hand uncertain requests to people instead of guessing.
Private beta Scoped pilots · Availability confirmed with your team
Why is shipment 4817 delayed?
Shipment 4817 is delayed because a port hold was recorded this morning. The current record does not include a confirmed release time, so SupportBot has prepared a cited update and flagged the timing question for human review.
Illustrative example · not live data
AIWF’s first commercial AI worker is designed for customer-support teams that already have documentation, recurring questions, and a clear escalation path.
Ground responses in the knowledge scope approved for the pilot.
Show the documents or records behind each answer.
Route uncertain or sensitive questions instead of forcing a response.
Bring SupportBot into a customer channel without changing the evidence and review model.
The normal path is evidence first. The exception path makes uncertainty visible instead of hiding it behind a confident answer.
A customer asks: “Why is shipment 4817 delayed?”
A customer asks: “Why is shipment 4817 delayed?”
SupportBot searches only the company knowledge assigned to the workflow.
Relevant shipment information and policy guidance are connected to the response.
A cited answer is delivered when the available evidence supports it.
Only approved outcomes should strengthen trusted business context; a conversation does not automatically become company memory.
Review conversations, inspect supporting sources, and follow escalations in the business console. Your team can check the context behind a response before deciding what happens next.
The pilot defines who owns reviews, which knowledge can be used, and which requests need a person.
Request a guided demo| Request | Evidence | Review |
|---|---|---|
| Why is shipment 4817 delayed?Timing needs confirmation | 2 sources | Needs review |
| Which documents are required?Approved documentation | 1 source | Answered |
| Can you approve an exception?Assigned to the support team | Policy | Escalated |
Illustrative example · not live data
Citations make a response reviewable. Verification checks help surface unsupported details or conflicting information. Neither is a guarantee that an answer is correct.
Agree which documents and approved answers belong in the pilot. External systems and live records require an explicitly configured integration.
Missing sources, policy exceptions, and sensitive requests need an agreed escalation owner.
Check representative questions, cited evidence, and escalations with your support team before increasing the scope.
The initial focus is logistics and similar businesses where customer questions depend on policies, shipment records, procedures, PDFs, and other operational knowledge.
Initial profile: 10–500 employees.
Customers ask about shipments, policies, service details, documents, and exceptions.
Your team already has approved documents, records, or answers that can define the pilot scope.
There is a person or workflow that should receive requests when the evidence is incomplete or sensitive.
A pilot is a practical evaluation of one support workflow, with scope and ownership agreed up front.
SupportBot is the primary commercial product and is currently offered through scoped private-beta pilots. Availability, onboarding, and commercial terms are agreed with each business.
A pilot begins with approved company documents, policies, procedures, or answers. Website sources and other integrations depend on the agreed deployment. The shipment example is illustrative and does not indicate an automatic live connection.
Define the escalation path during setup. Unsupported or sensitive requests should go to a person, with the available context, rather than being presented as a resolved answer.
No. The browser-local simulation uses illustrative responses. It does not connect to your company knowledge or create a production support request.
Define the knowledge scope, escalation path, and questions that matter most. Use the pilot to prove the operating model before expanding to more AI workers.