The intelligent era is here, but adopting AI is no longer the hard part. The hard part is turning that usage into measurable business impact
Glean's Work AI Index highlights that gap. While 87% of digital workers use AI and 75% say it makes them more productive, only 13% say their organization is performing significantly better as a result.
Closing that gap is the real challenge ahead. Getting AI into people's hands was step one. Step two is turning scattered wins into measurable, repeatable business impact. Companies have to identify which use cases are creating value, turn them into repeatable systems, and scale them without losing control as they spread across the business.
On August 26 and 27 at Fort Mason in San Francisco, we’re bringing customers, builders, admins, and AI leaders together at Glean:GO 2026. Over two days, we’ll focus on the decisions that turn AI from a promising tool into something the whole business runs on.
Here are five things you’ll learn about turning AI adoption into company-wide impact.
1. Learn how to prove AI is creating business value
An adoption dashboard can tell you 90% of employees logged in last week. It says nothing about whether the company is performing any better because of it. That second number is the one executives have started asking for, and it's the one most AI programs still can't produce.
The Outcomes track at Glean:GO is built for the people on the hook for it. “How to define and tell the story of AI ROI” will show how to connect AI investments to clear baselines, realistic targets, and KPIs that executives and finance care about.
“Escape pilot purgatory,” led by Tim Glomb, Co-Founder and Managing Partner of Enterprise AI Transformation at GleekOps, will examine the decisions that determine whether a pilot scales or stalls.
The “Inside AI Transformation” customer sessions will show how that work plays out inside real companies:
- Motive’s Head of People Analytics and Technology, Yuyan Sun, will share how the company rolled Glean out to 4,000 employees in six weeks, sustained adoption after launch, and now compares high and low adopters against business metrics like revenue growth and engineering velocity.
- Dell’s Director of AI Product Deployment and Management, Connor Bogin, will explain how the team sustained usage and connected adoption to outcomes like pipeline and attainment.
- Howard Hughes’ VP of Enterprise AI, Jonathan Goldberg, will trace how a 100-user pilot and a series of early wins led to a CFO mandate for a company-wide deployment as part of a plan to cut G&A expenses in half by the end of 2026.
This is the shift from metric disappointment to metric discipline — gathering enough evidence to decide what deserves more investment. Once a use case clears that bar, someone has to build it.
2. Discover what it takes to get an agent into production
Proving impact assumes you built something that can produce it. Most agents never get that far. They fail somewhere between the demo and production.
Closing that gap increasingly requires a forward-deployed engineering mindset: get close enough to the workflow to understand the real problem, then make the technical decisions needed to solve it. Sometimes the answer is an agent. Sometimes it’s better retrieval, a permission-aware action, or a more precise definition of the job itself.
The Builders track follows that work from definition through deployment. In “From AI use case to AI solution on Glean,” Janki Vansia, Senior Manager of Digital Experience at GoodRx, and Sean Hanna, People, Process and Technology Consultant at HubSpot, will show how to map the actual workflow and make the design choices that need to happen before a team starts building.
“Inside the Agent Development Lifecycle” provides a repeatable approach for building and scaling what works without creating a graveyard of one-off experiments.
In “Under the hood: Customer developer stories building with the Glean platform,” customers will demo working products and unpack the architecture behind them, including the APIs they chose, how they handled authentication and permissions, and what they built themselves.
Once an agent reaches production, the challenge changes. Every new action creates a boundary to define, and every model call, tool call, and reasoning loop adds to the bill.
3. Learn practical ways to govern AI and manage its cost
As AI takes on more complex work, two questions become unavoidable: Is it safe? And what is it costing us? IT and platform teams often end up owning both, usually with less visibility than they'd like.
The Admin track is designed to give them a more concrete way to evaluate each.
Security
“Five AI agent security assumptions we’re putting to the test” will challenge the ideas teams often accept at face value, including “permissions-aware means secure,” “read-only agents are low risk,” and “audit logs are enough.” Each assumption will be tested against a real scenario.
Then, in “Guardrails in action,” you’ll put those lessons into practice: scope an agent, identify its risk surface, write policies that match, and test the controls before deployment.
Cost
“How tokenomics ate the world” looks beyond model pricing to the architecture underneath the cost. Excess context, repeated tool calls, long reasoning loops, and poor model choices can all drive spend higher as agents take on more complex work.
The session will examine AI economics as a question of consumption and yield: how many tokens a system uses, what business value that consumption produces, and when a premium model earns its price.
The goal is to move quickly without losing sight of each system’s risk, cost, or expected value.
4. Get clear on the strategic calls AI leaders need to make
Those tradeoffs don't resolve on their own. Somebody at your company now owns AI. Maybe it's a new CAIO (Chief AI Officer). Maybe it's a CIO whose remit just expanded. Either way, someone has to decide where the technology is dependable, how work should change, and what the company needs to operate it at scale.
The AI Strategy track is built around those questions.
Where does AI break? “The borders of AI: The value of understanding where AI breaks” will examine the failure modes that polished demos tend to hide: writing that sounds competent but says little, analysis that appears rigorous but is quietly wrong, and long tasks where small errors compound. The session will help you decide where verification belongs, where a person should remain involved, and where human judgment still carries the decision.
How should work change? “How AI is really reshaping work: which jobs are changing and why” will explore which roles are being redefined and why stronger results may require redesigning the work itself.
What does scale require? General Motors expanded AI access from zero to 80,000 users in 90 days. In “From Rollout to Transformation: GM’s Operating Model for Enterprise AI,” Nikhil Sandilya, Director of AI at GM, will share how the company approached the harder phase that followed, including the business case, the prioritization of high-value use cases, and the operating model behind secure, company-wide transformation.
Together, these sessions address the judgment required to run AI as a company-wide discipline.
5. Understand the platform requirements for AI at scale
All of these decisions eventually become a platform requirement. If AI is going to move from scattered wins to a company-wide system, it needs the right context, governed agents, and a platform that can connect to the rest of the technology stack.
The Product track shows how Glean is building that foundation. “Understanding Glean Enterprise Context” explores how Glean models an organization’s people, content, workflows, and individual work patterns so AI can surface the right information and action at the right moment. “Glean Agents deep dive: Building, governing and deploying AI coworkers” will cover new capabilities for builders and admins, using departmental examples to show how agents can be applied across the business.
For teams making broader platform decisions, “Glean as a company-wide AI platform” will go deeper on the APIs, SDKs, embedded experiences, and gateways that connect Glean to the broader AI ecosystem.
Together, these sessions answer the practical question underneath every strategy discussion: what can your company build on now, and what should it plan around next?
A conference built for the work ahead
Five distinct tracks, hands-on workshops, certification sessions, and an Expert Lounge where you can bring a real problem and work it through with a Glean specialist instead of only hearing about someone else's.
- Admins can focus on governance, deployment, platform health, adoption, and cost.
- Builders can dig into workflows, architecture, APIs, and agent development.
- AI leaders can examine measurement, workforce change, strategy, and operating models.
Bring the team responsible for making AI real, and give each person the sessions, tools, and examples they need to succeed in this new intelligent era.
Save your spot
Enterprise AI is entering a crucial phase. The question is no longer whether employees will use it, but whether companies can turn that adoption into something measurable, governed, and worth scaling.
Glean:GO 2026 brings together the people doing that work. Join us at Fort Mason in San Francisco on August 26 and 27, or attend virtually, for two days of practical sessions, real customer stories, and hands-on work with the teams building and operating AI at scale.
Register for Glean:GO 2026 today!
