AI search
Customers get direct answers from your articles.
[ INTAKE · PRODUCT ]
Answers, acts, and knows when to stop
Docs help, but they are not enough when the answer depends on the user’s plan, settings, usage, or account state. With Actions, Intake can also call your API to resolve it.
Acts, not just answers
The agent calls your API to resolve, not just to explain
You define the endpoints and access level. The agent makes GET and POST requests to your product to check real state or run the next step.
AI resolution
Resolve repetitive tickets before they add team load
AI answers with documentation and real account state. If it cannot resolve, it escalates with context.
Human control
Escalate to a person with full context
When AI should not answer alone, you get the thread, history, account state, and what the agent tried.
Contextual docs
Docs are not enough when the answer depends on the account
The agent combines articles with plan, settings, usage, and state to give a specific answer.
A knowledge base connected to daily work
Publish answers, connect documentation, and let AI use trusted knowledge before escalating to the team.
Explore help centerConnected docs
The team replies with up-to-date and consistent knowledge.
Team editor
Create clear articles with images, blocks, and useful sections.
Install the agent where the problem happens
Messenger widget
Embedded widget inside your SaaS to resolve where the user gets blocked.
Email
Turn repetitive emails into tickets with visible account state.
Slack
Handle Slack questions with account context before replying.
An inbox for tickets that need context
Every conversation arrives with history, plan, settings, usage, and what the agent already checked so you resolve faster.
AI resolution
AI answers with docs and real account state to reduce repetitive tickets.
Customer context
The team sees history, plan, previous conversations, and relevant signals without opening other tools.
Account data
Plan, settings, usage, and state next to every conversation.
Clear handoff
Notes and context when a person needs to step in.
Deflection
Measure which tickets did not reach the team.
Multilingual
Help the team reply to customers in different languages.
Privacy, control, and human oversight
The agent is designed to work with customer data in a controlled way, with secure access and human decisions when they matter.
GDPR-ready
Secure access
Data control
Frequently asked questions
Does AI answer with our information?
Yes. It answers using documentation and real user context: plan, settings, usage, and account state.
What are Actions?
Actions are GET and POST calls the agent makes to your API within the permissions you define, to check real state or run a change instead of just describing it.
What is the agent?
The agent is an AI support agent embedded in your SaaS. It reads real account context and answers repeated questions before they add load to the team.
Does the agent replace the team?
No. The agent reduces repeated work and escalates the cases where a person adds judgment.
Can we use our own documentation?
Yes, but it does not stop there. The agent combines docs with plan, settings, usage, and real account state.
How do we get started?
Most teams start with the repetitive questions creating the most load and the account data needed to resolve them.
