Sequel, a webinar platform that runs events on a company’s own website, made its AI agents generally available to every customer on 2 November. Its pricing page says agent usage is free on all plans until 1 March 2027. After that, Sequel will charge for “the AI work Sequel completes, measured in actions”. It has not published a price per action.
So a small marketing team gets about three and a half months to test the agents on real events, but no figure to budget against yet.
What the three starting agents do
Build plans and prepares events. It suggests topics and timing from the team’s own attendance history and can clone an event from a template. Its Event Check runs 72 and 24 hours before each session, tests registration, emails, the landing page and CRM sync, and posts only the problems to Slack, according to the launch post.
Scale works after the session. It cleans the transcript, cuts clips and drafts a blog post, a LinkedIn post per speaker and a newsletter blurb. Convert drafts a follow-up for each attendee built around the question they asked, and briefs account owners in Slack.
Teams can also describe a custom agent in plain language, or use Sequel’s MCP server (MCP is the open standard that lets an AI assistant read and act in another app). The agents page says it works with Claude, Cursor and other MCP tools. Our guide to what an MCP connection can reach covers what to check first.
Nothing goes out without a person
The agents read and draft on their own. Sequel says they cannot send email, change events, publish or push leads to a CRM until someone approves. That is the right default, and it is the same approach as Box’s plain-language workflow builder, where a person approves changes.
The cost is review time. A sample screen on Sequel’s own page shows 486 follow-up drafts waiting for approval after one event. Someone has to read a sample before approving the batch.
On data, Sequel’s FAQ says the agents use OpenAI’s GPT models through the business API with zero data retention, and that customer data does not train models. Attendee profiles combine site visits with what people asked in a session, so EU teams should still run the usual checks before customer data goes into an AI tool.
What it costs and what setup takes
The agents come with the platform. The Growth plan, for sessions of up to 500 attendees, is listed at $950 a month, or $833 a month billed annually. Business is priced on attendee volume and Enterprise on a custom contract. Every plan starts with a sales demo.
Sequel says setup takes one session. Agents work best with Slack connected, a brand kit and its tracking pixel on your site. The FAQ adds that not every action works the same way for live, simulive and on-demand events.
Read the results as vendor numbers
Sequel says Unit21 now gets 19 assets from one 60-minute webinar with about 3 hours of review, against an estimated 29 hours by hand. It also cites $14 million in attributed pipeline at ElevenLabs. These figures come from Sequel and its customers (the ElevenLabs one is also in the Business Wire release), and we could not confirm them. The Marketing Exploit notes they reflect sophisticated setups.
What to do this week
If you use Sequel, turn on Event Check for your next webinar, since it only reports problems. Then run Scale on one recorded session and time the review. Count the agent actions a typical event uses, so you can price it once Sequel publishes a rate.
If you do not, list the post-webinar tasks your team skips, and check whether your current platform already covers them.
Sources
Get the weekly email
The week's most useful AI news, a tutorial and an interview.




