First Impressions of Expedia’s Glean AI Enterprise Portal

These notes are a bit dated, but I'm sharing them as early learnings, while I work on articulating more recent examples of using AI tools to amplify my role, cut through noise, and support my team.

When Glean was introduced, I wanted to understand how I might leverage it in my role as a design manager and leader, to inform better strategy, guide delivery, support colleagues, and enhance design workflows.

At first, Glean looked like any other work portal I had seen over the last 15 years: a familiar search bar, company articles, and links to recent documents. However, after making a few search queries and exploring the agents, it quickly felt less like a tool and more like a company librarian, personal assistant, or even a colleague who had been around for a while.

What stood out wasn't just access to information; it was how quickly it helped bring focus to what needed attention. 

Search vs Chat: Different Modes of Thinking 

I naturally gravitated toward the Chat interface because of its conversational flow. It felt intuitive and aligned with how I already interact with AI tools.

That said, the Search experience proved to be more comprehensive when I needed to:

  • Locate documentation across multiple systems and departments

  • Browse internal resources, locate support and access training

  • Consolidate information from different teams

Together, they created a useful balance:

  • Search for breadth and discovery

  • Chat for synthesis and interaction 

Where It Became Valuable: Reducing Cognitive Load

 The biggest shift came when I started using agents, particularly the pre-built Intelligent Reminders agent.

As a design manager, a lot of my role involves:

  • Tracking conversations across Slack, Jira, Figma, and meetings

  • Informing the status of items and decisions

  • Supporting multiple designers across different workstreams

That mental load adds up quickly.

Instead of manually checking everything or relying on others to raise issues, the agent surfaced a consolidated view of where I was needed.

Some examples:

  • It reminded me that I had posted in Slack that I'd miss a leads meeting, but hadn't received responses to the notes I had left in advance.

  • It flagged Jira tickets across my teams that were blocked or had open questions.

  • It nudged me to update assets for upcoming ceremonies and design critique sessions.

  • It highlighted unresolved Figma threads related to a high-profile release (including missing testing details and timelines i was chasing).

  • It surfaced conversations where designers had tagged me to review feasibility discussions or design system proposals.

  • It even picked up smaller but important things, like confirming attendance for a team baby shower. 

What made this useful wasn't just visibility, but helping to focus on what matters by surfacing:

  • What might have been missed?

  • What actually needs my attention?

  • Where can I unblock others?

Using AI Beyond Retrieval: Building My Own Agents

I also started experimenting with creating my own agents.

News Filter Agent

I built a simple agent to filter news across topics I care about, starting with global news and expanding into tech, travel, and design.

It helped:

  • Reduce noise

  • Surface relevant trends

  • Keep me strategically informed 

Mapping Feedback to Role Profiles (+ “Superpowers”)

One experiment I'm particularly interested in is mapping feedback to role profiles and identifying what someone's "superpower" might be.

Feedback is often:

  • Fragmented

  • Subjective

  • Hard to translate into clear development actions

This agent helps:

  • Map feedback from peers, managers, and self-assessments

  • Identify patterns in strengths and gaps

  • Make growth conversations more structured and grounded

It shifts AI from being a summarisation tool to something closer to a coaching assistant. 

Where This Could Go Next

There's still room to evolve, from better information to better systems for thinking and supporting my teams. Some areas I'm exploring:

  • How to evolve from summaries to decision support (Risks, opportunities, recommendations)

  • Identifying gaps in coverage (what the system can't see)

  • Connecting tools like Glean, Claude, Cursor, and n8n into more complete workflows 

Final Thoughts

Glean is relatively new to Expedia, and in a few months, these examples will likely feel basic. I am excited to share how things evolve. However, I am discovering that AI isn't just about speed or automation. For me, it's quickly adding value in reducing noise, improving clarity, and helping me show up more intentionally for my team.

---

Senior Manager, User Experience Design at Expedia Group.

Previous
Previous

Turning feedback into focus: using AI to fill gaps and amplify strengths

Next
Next

AI Hackathon Submission: “Family travel is constraint-led, not casually exploratory”