Skip to content
Agents tracked: 501 Releases (7d): 777 Agent status: no reported issuesSince last limit reset: Claude 10d · Codex 8d For AI agents: Markdown Updated Nov 21, 2026
Case study SubstackResults as reported by Decagon

How Substack automated Tier 1 reader and publisher support with Decagon

Substack used Decagon's AI agent to handle repetitive support requests such as cancellations and email imports. Decagon reports that most inquiries are now resolved without a human.

The problem

Agents handled a large volume of repetitive Tier 1 requests by hand. Substack had little customer segmentation, routing or analytics on support trends. The team also wanted time to work proactively with selected publishers.

How they did it

  1. Deployed the AI agent on incoming user inquiries.
  2. Added custom segmentation filters and tagging so different users get tailored support.
  3. Connected the agent to internal APIs so it can carry out multi-step actions such as refunds and cancellations.
  4. Linked it to existing tools for logging bugs and feature requests, and used its Voice of Customer analytics.

Results as reported by Decagon

  • Decagon reports that the agent resolves more than 90% of user questions without human intervention.
  • Qualitatively, the source says support capacity grew without adding headcount, and human agents moved to higher-value publisher support.

These figures are as reported by Decagon in Decagon customer success story ↗. Decagon sells the product used here. AgentGid did not measure or independently verify them.

Takeaway. Connect the agent to backend actions like refunds and cancellations, not just FAQ content, if you want it to fully resolve Tier 1 tickets.
AgentGid's take

This fits a support team with enough ticket volume to justify engineering time, since the setup depends on internal APIs for refunds and cancellations plus custom segmentation. Decagon has no public pricing (Custom, via sales), and the 90%+ resolution figure is vendor-reported, so get a quote and a pilot scope first. Rasa (Free) is a higher-ranked open-source option if you can build it yourself.

The agent used here

Similar use cases

Guides