
How Zapier tackled AI sprawl and doubled ticket deflection rates with Glean
Introduction
Zapier is an AI orchestration platform built for teams at organizations of every size, helping teams connect with over 9,000 apps to automate the most tedious parts of the workday. It turns complex integrations into simple, point-and-click workflows that handle everything from lead routing and data synchronization to customer conversations, all without requiring users to write a line of code. Since 2012, millions of users—from scrappy startups to Fortune 500 companies—have trusted Zapier to streamline their operations, reduce errors, and accelerate growth through intelligent automation.
Avant Glean
At Zapier, giving people the freedom to experiment with and adopt the tools they liked has enabled everyone to do their best work. However, that often meant information was scattered across sources, making it challenging to settle on a single system of record.
Consolidating those tools was one option, but it meant losing resources that teams relied on for important workflows, as well as taking on what would have been a massive, time-consuming project. Zapier needed a way to consolidate and surface the right information at the right time, without asking people to change how they worked or give up the tools that fit their needs.
That’s why they turned to Glean.

Avec Glean
Unifying knowledge across teams and apps
With information at Zapier living across many internal tools, teams like Revenue Operations, who field a constant stream of questions from across the company, turned employees to Glean to find answers on their own.
Utilizing Glean’s Chrome extension to ensure enterprise context was at every employee’s fingertips, they quickly saw adoption rise. “Glean is a tool that speaks for itself once you start using it,” said Elise Dunham, Knowledge Manager.
Ticket deflection powered by Glean
Zapier launched ZapIT, a Slack-native helpdesk built on Zapier’s own platform, with Glean powering AI ticket deflection and knowledge search. With Glean providing the layer supplying enterprise context, including Slack history and company documentation, users could get answers directly in the Slack thread where the ticket was created.
This more than doubled successful ticket deflections, taking an average of 30 minutes to close. 98% of tickets opened at Zapier are now answered with Glean, with a 99% relevance rating. Employees are able to get answers much faster, freeing up their IT team to prioritize more complex organizational initiatives.

Bringing context into agent harnesses via MCP
As Zapier employees started working directly inside agent harnesses like Cursor, Claude Code, ChatGPT, and Codex, setting up MCP servers that connected to Glean proved invaluable. Employees can now pull unified context from Glean directly into the tools they are already working in.
Agents Driving Business Outcomes
Zapier's Go-to-Market team developed a sales agent that pulls together the full context of a customer before a meeting, spanning product usage, back-and-forth in customer Slack channels, and internal notes from the account relationship.
With a single clear snapshot of a customer’s current state and recommendations on how best to engage them, GTM teams are more prepared for customer conversations, driving faster time to value and immediate impact. The Agent’s impact extends beyond sales as well - Executives, Marketers, and Product Managers also utilize it as a resource to prepare customer briefs before meetings and review cycles.
Zapier’s Customer Advocacy team maintains a structured library of more than 700 customer stories, quotes, metrics, and reference assets, enriched with metadata such as industry, use case, persona, product, freshness, and usage restrictions. Glean gives Zapier teams a natural-language interface to that knowledge, helping GTM teams find relevant proof without needing to know where the underlying asset lives or how the database is structured.
The goal is not simply to store more content, but to make customer proof available whenever a salesperson, marketer, or customer-facing teammate needs it, with safeguards for freshness, marketing clearance, and customer-naming restrictions.
Explore more customer stories
Work AI qui fonctionne.














