The questions every Salesforce team should ask about AI at Dreamforce 2026

AI will be the subject of most conversations at Dreamforce this year, whether that’s in a session, at a booth, or standing around with people between talks. But cutting through the buzzwords and hype can be hard, and you don’t want to get back to the office with no clearer insights on how to actually implement AI most effectively for your team. So instead of another list of sessions to attend, this post will walk through three key questions to help you come away from Dreamforce with answers specific to your own org, not just AI hype.

If you’d rather skip ahead and talk this through with someone in person at Dreamforce, you can grab dedicated time with the DevOps and agentic change experts at Gearset: book your 1-to-1 here. Otherwise, here are the three questions worth raising with anyone you talk to this year.

How do you review changes when AI is exponentially increasing throughput?

Before AI, the number of changes moving through your Salesforce release process was limited by how fast a person could write them. AI-assisted development removes that limit. For many teams, this means their review process is having to handle a significantly higher volume of change than it was originally built for.

Teams relying on a manual review process tend to end up either with an overwhelming backlog or being less thorough to keep pace. Teams that have been able to accommodate the speed of AI-assisted development, without lowering security or quality standards, have typically invested in a robust test automation suite.

At Dreamforce, it’s worth asking people how their review process has changed to accommodate AI: if their review process hasn’t changed, how are they keeping pace or have they had to make compromises? And if they have invested in automation, how have they configured their test suite and what advice do they have?

Do you know what’s changing in your org and can you fix it when it breaks?

Another challenge with the volume of changes produced by AI is visibility. As more changes get proposed and shipped by agents, it gets harder to keep track of what’s in flight and what’s already gone into production.

That gap shows up in two ways. Firstly, it’s harder to produce a clear audit trail of what changed, when, and who reviewed it, which can cause compliance issues. And secondly, problem-solving gets slower, because when something does go wrong, tracing it back to the change that caused it takes longer if you didn’t have a clear view of what was moving through the pipeline in the first place.

The good news: this is solvable with the right Salesforce DevOps tooling. Here are a few questions worth asking other Salesforce teams to understand how they’re configuring their visibility practices: How do you currently track AI-assisted changes that are in flight and are awaiting review? If an incident happened today, how do you trace it back to the change that caused it? And when your auditors ask what changed and who approved it, what documentation or tracking practices do you have in place to be able to answer that question?

How does the role of a team lead change in the era of AI?

A big part of a team lead’s job is defining team strategy and goals, and allocating work accordingly. AI now needs to be a central part of that planning and allocation process.

First, a team lead needs to accurately understand where the team actually sits on the AI maturity scale today and where it needs to get to. Most teams currently use AI as a copilot but are striving for a virtual teammate, with defined responsibilities, governance and compliance built-in, and with deep understanding of your specific org.

The team lead then needs to outline a strategy to reach that maturity goal and manage team capacity and work to make that achievable. Getting to a virtual-teammate model takes deliberate investment: time for people to build review skills, time to establish DevOps guardrails, time to course-correct early mistakes. None of that happens if capacity planning only ever asks who’s free for the next ticket.

So at Dreamforce this year, ask: does your team have a defined vision of your ideal AI setup? What are your key tactics for getting there? And how are you managing your team's capacity to make that vision possible?

Bringing this into your conversations at Dreamforce

That’s a lot to carry into a conference before you’ve even checked into your hotel, so it’s worth pointing out: you’re not expected to have all three solved before you go. In this year’s State of Salesforce DevOps report, 76% of Salesforce professionals told us they sit somewhere in the middle on AI, neither convinced it will do everything nor dismissing it as hype. Most people are still working this out too, so you’re not behind if you don’t have it all figured out.

What’s genuinely useful is spending some time before Dreamforce working out which of these three questions is your team’s actual weak point right now. A specific question tends to get a specific, useful answer in conversation. A vague one gets a generic response you could have read in any blog post.

How to find answers and next steps for your team

Two sessions at Dreamforce this year are directly related to these questions, if you want to hear how Gearset, the team behind DevOps Launchpad, is thinking about them.

Onboard your virtual teammate. Assign it half your tickets.

Covers the first two questions above from the technical delivery side: what it actually takes to trust an agent with real tickets in your Salesforce org.

Your evolved role: human skills for admins managing agents

Covers the third question: what the admin or team lead role looks like once part of the work is being picked up by an agent.

While these sessions can tell you how other people are thinking about this, they can’t tell you whether your review process has a specific gap in it, or whether your team needs to rethink what gets delegated, etc. That’s the kind of conversation worth having in person, which is why we’re running dedicated 1-to-1 sessions throughout Dreamforce.

For over ten years, Gearset has helped teams deliver Salesforce changes quickly, safely, and robustly. Now we’re answering the same challenges in the agentic era. So whether you’re focused on speed, security, governance, or getting started with AI at all, bring your Salesforce setup, and leave with next steps you can actually apply when you’re back in the office after Dreamforce.