Building an AI agent in Pod should feel simple. The goal is not to create โan agent that does everything.โ Itโs to build an agent around a specific workflow moment, with a clear job to do and an output your team can actually use. That mirrors the approach emphasized in Podโs best-practices guide: agents work best when they are focused, actionable, easy to scan, and easy to iterate.
Introduction: Why build a custom AI agent?
Pod AI Agents are most useful when they help a seller or manager make a decision in a real moment of work.
Examples:
โIโm preparing for a deal review.โ
โI need to know whether this deal is actually healthy.โ
โI want to spot stakeholder gaps before the next call.โ
โI need a mutual action plan after a PoC discussion.โ
A well-built agent creates consistency across your team and saves time on repeat analysis. A poorly built one usually becomes too broad, too verbose, or too generic to be useful. Podโs own guidance recommends starting narrow and building around a concrete workflow moment rather than trying to analyze everything at once.
Where to find the Agent Builder in Pod
You can access the AI Agent Build in two places:
A. In your Settings page
Access your Account Settings
Select the "Agents" section
You will find your agent triggers (see Autonomous Triggers article for more details)
Click on the "Manage Agents" at the top right of the screen.
B. Directly in Account 360ห
Access any record in Account 360ห
In the Deal Coach tab, click on the "Explore Agents" in the AI Agent input box
In the expanded AI Agent page, click on the "puzzle" button to open the modal.
What types of Agent Builder can you find?
Inside the agent builder modal, youโll find three types of agents:
Personal Agents: These are agents you create for yourself.
Org-Wide Agents: These are shared across the organization.
Pod Pre-Built Agents: These are Podโs out-of-the-box agents.
The Org-Wide Agents can only be built by workspace admins.
How to build an AI Agent?
When you have the Agent Build modal opened, you can now build a new agent.
Open the Agent Builder.
Choose the agent type: Personal Agent or Org-Wide Agent (admins only).
Give the agent a clear name that reflects its purpose (e.g., Deal Health Check or Discovery Gaps Finder).
Add a short description explaining what the agent analyzes and what outcome it produces.
Write the prompt defining what the agent should evaluate using the available deal context.
Define the response format so the output is structured and easy to scan (summary, risks, next actions, etc.).
Test and refine the agent, then save it and start using it in your deal workflows.
For more details, review our AI Agent Best Practices Guide to know how to build great agenti prompts.
๐ก Need help? Send us a message via the in-app chat or email us at [email protected].
๐ค Want to talk to someone? Book a session with one of our specialists!