Productlane Agent
Instant answers out of your docs, roadmap, and changelog.
Productlane Agent
The Productlane Agent answers customers and resolves conversations across your widget, live chat, and public docs. It is grounded in your help center, roadmap, and changelog, and it replies in the customer's language.
When a conversation turns out to be a bug or a feature request, it can file a Linear issue. When it cannot resolve something, it hands the thread to your team.
The Agent needs the Starter plan or above, and it's on by default for new workspaces.
Where to find it
Settings > AI > Productlane Agent, at /settings/ai/agent. A live widget preview sits beside the settings, so you see the Agent as your customers will.
Set it up
Turn on the Productlane Agent.
Edit the Welcome Message and the three Suggestions to match your product's voice. With several languages enabled, pick each one in the selector and translate them.
Add Workspace guidance: tone, policy, and product context the Agent should follow.
With Linear connected and a default team set, turn on Create Linear issues from agent chats so bug and feature-request chats become issues.
To have bugs fixed automatically, connect Cursor, then turn on Auto-assign Productlane Agent to bugs, and Exclude manually created issues if you want that narrowed further.
Turn on Generate auto-replies with AI to add an AI section to your email auto-replies.
Everything saves as you change it.
What grounds its answers
Your help center docs, roadmap, and changelog by default. That includes link entries in your docs sidebar, pages you host elsewhere that the Agent reads and cites with their real URL.
Connecting GitHub maps every customer-facing screen in your linked repository into agent-only articles, and the Agent grounds its answers about the product in those. It never reads your repository directly. Connecting PostgreSQL lets whitelisted read-only SQL inform answers. On email auto-replies it also reads files and pasted screenshots on the customer's message, up to five per message.
Anonymous visitors
The Agent answers a visitor with no known email the same way it answers an identified one, for as many turns as the conversation runs. It never interrupts a reply to ask for an address.
An address only comes up when the visitor asks for a human. See below.
Asking for a human
A customer can ask for a person at any point. The Agent recognises "talk to a human" or "connect me with support" and shows a Talk with a human button rather than pressing on.
If the widget does not already know the visitor's email, clicking the button asks for one first, since the team needs an address to reply to. The handoff runs as soon as they add it, and with Live Chat enabled the conversation moves into Live Chat on the same thread, so the customer keeps reading one continuous conversation. Once someone has asked for a human, the Agent stops replying there, Live Chat or not.
What a handoff does
A thread can land in your team's hands two ways, and they behave differently.
The customer asks for a person. The thread comes back to your inbox, marked unread and high priority, and it stays that way:
It stays open. Productlane never marks it Done on its own again, so it keeps appearing in your open views until a teammate picks it up.
The timeline says so. The activity timeline logs a "Customer asked to talk with a human" entry.
Your team hears about it. Turn on Productlane Agent handoffs for Slack and email under Settings > Notifications to be paged the moment one happens.
The Agent decides on its own that the conversation needs a follow-up, without the customer asking. This reopens a closed thread and flags it for your team, but it does not freeze the thread the way a customer request does: if the conversation goes quiet afterward, Productlane can still mark it resolved automatically later. The timeline logs this as "Productlane Agent handed this conversation to the team".
Controlling handoff
There is no separate switch for handoff. The Agent always hands a thread to your team in the two cases above: when a customer asks for a person, and when it decides on its own that it cannot resolve the request. The only way to stop it from handing off is to turn off the Productlane Agent itself, which also stops it from answering.
Use Workspace guidance to shape when and how it escalates. Write in the topics that should always go to a person and the tone it should use when it hands a thread over, the same field you use for general tone and policy.
Whether an escalated thread moves to Live Chat or stays in the Agent's chat depends on your Live Chat setting, not on the Agent:
With Live Chat on, the thread moves into Live Chat on the same conversation, and the customer sees a confirmation message.
With Live Chat off, the thread still comes back to your inbox as unread and high priority, but the customer stays in the AI chat interface and never sees a separate handoff notice or a response-time message.
The Agent cannot send an email instead of handing off. See the Live Chat article for how that toggle changes what customers see.
Reading its activity
When the Agent runs three or more actions on a thread inside three minutes, the timeline collapses them into one entry, such as "AI Agent ran 7 actions". Expand it for the full list. Failures stay visible even when collapsed, for example "AI Agent ran 5 actions, 2 failed".
A routine or a docs proposal that gets skipped because the AI spend limit is reached says so directly in the timeline, with a Raise the limit link to Settings > AI > Usage and credits.
Limits
Each confident, non-escalated resolution counts against your plan's included resolutions for the month, whether the Agent resolves it directly in the widget or live chat, or an AI auto-reply resolves it over email, Slack, or Microsoft Teams without escalating. A message a teammate sends while testing the Agent in AI chat never counts as a resolution, on a trial or on any paid plan.
Once those are used up, resolutions stop unless pay-per-use is on. Caps and usage live under Settings > AI > Usage. When the spend limit stops the Agent, Settings > AI > Productlane Agent shows a Paused badge next to the on/off switch even though the switch itself stays on.
Each visitor is capped at 150 chat messages a minute per IP address and 50 a minute per session, and a single message can run to 4,000 characters. At 50 messages, the widget asks the visitor to start a new chat.
If the Agent cannot reach its docs, roadmap, or changelog search when someone asks a question, it says so and offers to connect them with your team rather than guessing.
Was this helpful?