Skip to content
GuideAgent setupIntermediate

Configuring your agents

The hub is one card per agent, and opening a card is not configuring it. What "configured" actually means, and the order that makes it go quickly.

RDRevTech DeliveryAgent delivery team 2 min

Product walkthrough

See it in RevTech

See the controls that define what each agent can do, the standards it must follow, and the usage RevOps can observe.

See the controls that define what each agent can do, the standards it must follow, and the usage RevOps can observe.
Part 1 of 4

One card per agent

The AI Agents hub is the entry point for every agent available to you, one card each, and each card opens that agent's own configuration page. The list only ever shows the agents your role and workspace have access to, so a card you expected and cannot see is an access question rather than a missing capability.

Onboarding counts agent setup as done once you have configured at least one of every agent type available here. Which raises the question of what configured means, and the answer is more demanding than it looks.

The RevTech AI Agents hub, where each agent is configured and governed.
The AI Agents hub: every agent RevTech runs, what it is capable of, and where its configuration and guardrails are set. Screenshot of the RevTech application; sample data.
The AI Agents hub: one card per agent your role can reach, each opening that agent's own configuration page.
Part 2 of 4

Configured means work saved

Configuration is real work saved on the agent's own page: a play, a lead flow, a campaign, an analysis, an SLA rule, or a messaging sequence. Simply opening a page does not count, and the product is deliberate about not letting it count.

That is worth stating plainly because the alternative design — marking an agent configured on first visit — produces a workspace that reports itself as set up and does nothing. An agent with no rules, no plays and no flows has nothing to act on. It is not idle because it is broken; it is idle because nobody has told it what the business does.

Part 3 of 4

Artifacts are the shared input

Every agent draws on the same governed artifacts — your ICP, product material, messaging context. That library is not per-agent, and it is not filled in for you. Whatever is in it is what all of them have to work from.

Which makes artifacts the highest-leverage thing on this page even though it is not one of the cards. An agent configured against a thin library produces thin output, and the instinct in that situation is to re-tune the agent, which is the wrong end of the problem.

Part 4 of 4

A workable order

You can configure agents in any order and onboarding will track what is still outstanding, updating as soon as you save. That said, some orders are less painful than others.

Get the artifact library to a usable state first, since everything reads from it. Then configure whichever agent supports the motion your team is already running: a lead flow if inbound is the priority, an SLA rule set if CRM hygiene is. Starting with the agent that supports a motion nobody runs yet produces a configuration nobody validates.

What to take away

  • Configured means real work saved — a play, a flow, a campaign, an analysis, an SLA rule, a sequence. Opening a page does not count.
  • The hub only shows agents your role and workspace have access to. A missing card is a permission, not a gap.
  • Agents work from the artifacts you maintain. An unconfigured artifact library caps what every agent can produce.
  • Order does not matter. Configure in whatever sequence suits you; onboarding tracks what is outstanding.
Demo

Try the demo.

See agents carry the repeatable work of GTM across sales, marketing, customer success, and RevOps. Every action prepared, reviewed, and recorded. Fictional data, real product.

Explore the demo