Back office ops · Production

Zillow's adoption of Glean builds an AI-forward culture with 1.5+ hours saved per employee weekly and over 3,400 agents created

The problem

Zillow's permanent shift to remote work scattered information across multiple applications and eliminated the informal office connections that previously facilitated knowledge sharing, leaving employees unable to find what they needed and creating significant discoverability challenges.

Workflow diagram · grounded in source
1
Remote work triggers search need
trigger
“employees struggled to find what they needed”
2
Glean searches unified data corpus
ai_action
“Glean connects over 30 data sources and 138 million documents, serving as a single source of truth”
3
Slack agents auto-handle HR questions
ai_action
“Slack support agents that automatically handle employee questions by accessing existing documentation, employee insights, and people analytics”
4
Ask Z Agent democratizes research
ai_action
“Ask Z Agent: Zillow's UX Research Team created an agent that indexes and democratizes access to customer research and insights. Now, employees across Zillow can easily utilize data-driven insights as they make decisions.”
5
Self Review agent synthesizes impact
ai_action
“Zillow created an agent that helped employees synthesize impact, and in doing so, drove a 7% increase in employee review participation in its first review cycle, run almost 10,000 times”
6
Incident timeline document created
output
“This Glean agent creates a first-pass of a postmortem document, organizing all communications and troubleshooting across multiple platforms into a single timeline, and even proposing action items”
7
MCP integrates context into coding tools
integration
“engineers connect Glean to AI coding assistants like Cursor and Claude Code, allowing them to pull context from design documents, JIRA tickets, and existing GitHub code patterns”
Reported outcome

Glean saves each Zillow employee 1.5 hours per week on information searching, achieved over 80% adoption across the company, and enabled the creation of more than 3,400 specialized agents across critical workflows.

Reported metrics
Search time saved per employee weekly1.5+ hours
Employee adoption rateover 80%
Glean agents created totalover 3,400
Agents created in first 6 weeks500
Show all 6 reported metrics
search time saved per employee weekly1.5+ hours
employee adoption rateover 80%
Glean agents created totalover 3,400
agents created in first 6 weeks500
employee review participation increase7%
Self Review agent runsalmost 10,000
Reported stack
GleanDatabricks GenieSlackCursorClaude CodeJIRAGitHub
Source
https://www.glean.com/resources/customer-stories/zillow
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Glean saves each Zillow employee 1.5 hours per week on information searching, achieved over 80% adoption across the company, and enabled the creation of more than 3,400 specialized agents across critical workflows.

What tools did this team use?

Glean, Databricks Genie, Slack, Cursor, Claude Code, JIRA, GitHub.

What results were reported?

Search time saved per employee weekly: 1.5+ hours; Employee adoption rate: over 80%; Glean agents created total: over 3,400; Agents created in first 6 weeks: 500 (source-reported, not independently verified).

How is this back office ops AI workflow structured?

Remote work triggers search need → Glean searches unified data corpus → Slack agents auto-handle HR questions → Ask Z Agent democratizes research → Self Review agent synthesizes impact → Incident timeline document created → MCP integrates context into coding tools.

Related back office ops cases