> ## Documentation Index
> Fetch the complete documentation index at: https://docs.introw.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI-native

> Agents do the mechanical work across the partner lifecycle - registration, follow-ups, content drafting, coaching, and support - while humans stay in the loop where judgment matters.

> Introw is AI-native, not AI-bolted-on. Agents take on the busywork so a small team can run a large program, and every agent is scoped by permissions and works both inside Introw and from the partner's own AI assistant.

## The problem

Partner programs scale headcount linearly with partners. More partners means more manual data entry, more follow-ups, more content requests, and more support tickets - until the team, not the market, is the bottleneck. Most PRMs answered this by bolting a chatbot onto the side: it can read a help article, but it can't safely take an action, so the real work still lands on a person.

* **Busywork eats the team.** Data entry and chasing updates crowd out the relationship building that actually grows the channel.
* **Support doesn't scale.** Every partner question routes to a person, in one time zone, in one language.
* **Bolted-on AI can read but not act.** A chatbot that can't touch the CRM safely just adds another dead end.

## How Introw does it

Agents work across the whole lifecycle, and they act - they don't just answer:

* **They take scoped actions.** A deal registers itself from a Slack message; a lead is captured, deduplicated, and written to the CRM - not handed back as a suggestion.
* **They coach and catch problems.** Deals get stage-specific selling guidance, and channel conflict is flagged the moment an overlap appears instead of weeks later.
* **They support partners 24/7.** Partners get an answer in their own language, grounded in your content library, at any hour.
* **They run where partners already are.** Over MCP, agents reach Introw from the partner's own Claude, ChatGPT, or Gemini - not only inside your app.

Agents are governed. Data access follows the signed-in user's partner scope, knowledge is scoped to the right context, and product entitlements gate what each agent can do - so humans stay in the loop exactly where judgment matters.

## What it looks like

A partner messages your shared Slack channel: "registering Acme Corp, \~40 seats, closing next quarter." The agent extracts the details, checks for an existing or conflicting deal, creates the opportunity in your CRM, and replies with the registration status - in under two minutes, with no one on your team touching it. If something needs a human call, the agent routes it for approval instead of guessing.

## Proof

<CardGroup cols={2}>
  <Card title="AI agents" icon="robot" href="/features/ai">
    Partner support and deal coaching, powered by your content.
  </Card>

  <Card title="Deal coaching" icon="chart-line" href="/features/ai/deal-coaching">
    Stage-specific selling guidance on every deal.
  </Card>

  <Card title="Channel conflict" icon="triangle-exclamation" href="/features/ai/channel-conflict">
    Overlaps caught the moment they happen.
  </Card>

  <Card title="MCP" icon="plug" href="/features/developer/mcp">
    Reach Introw from any MCP-connected AI client.
  </Card>
</CardGroup>

## The impact

* **Run a large program with a small team** - agents absorb the mechanical work so people focus on relationships.
* **24/7, multilingual support** - partners get answers without waiting on a person or a time zone.
* **Faster, cleaner registration** - deals enter the CRM in minutes, deduplicated and conflict-checked.
* **Safe by design** - agents act within permissions, with humans in the loop on anything that needs judgment.

## Common questions

<AccordionGroup>
  <Accordion title="How is this different from a bolted-on chatbot?">
    A bolted-on chatbot can read and reply. Introw's agents take scoped actions - registering deals, capturing leads, writing to the CRM, coaching deals - within the signed-in user's permissions, so they do the work instead of pointing at it.
  </Accordion>

  <Accordion title="What keeps the agents safe?">
    Governance. Every agent respects the user's partner scope, knowledge is scoped to the right context, and product entitlements gate what each agent is allowed to do. Anything needing judgment is routed to a human.
  </Accordion>

  <Accordion title="Do partners need to use Introw's interface to get the AI?">
    No. Agents run wherever partners already are, including their own Claude, ChatGPT, or Gemini over MCP, as well as Slack, Teams, and email.
  </Accordion>
</AccordionGroup>

## Keep exploring

<CardGroup cols={2}>
  <Card title="Previous: Low total cost of ownership" icon="arrow-left" href="/why/low-tco" cta="Back">
    The partnerships team configures everything, no engineering.
  </Card>

  <Card title="Next: Time to value" icon="arrow-right" href="/why/time-to-value" cta="Continue">
    Setup is configuration, not construction.
  </Card>
</CardGroup>