Braze used its Forge conference in Las Vegas this week to unveil new AI capabilities built for marketers, alongside a deepened strategic partnership with AWS.
Speaking to CX Today ahead of the event, George Khachatryan, VP of AI Decisioning Product at Braze, walked through what each launch actually changes for marketers, and why he believes AI is now less about replacing skill sets and more about collapsing them into one.
Decisioning Studio Go: 1:1 Personalization Without the Data Science Team
AI can help marketers to decide on which messages hit harder, and which campaigns will be most effective. But there's a catch. Setting up these systems required real technical investment: a data engineer on the customer's side and a data scientist from a vendor.
Braze's newly announced Decisioning Studio Go is built to remove that barrier entirely, letting marketers deploy a self-service AI decisioning agent that tests message variants, timing, day, and frequency for each individual customer, without involving a technical team at all.
"You need to have a team of data engineers, or at least a data engineer on your side. So that's a good fit for your highest value journeys. But what about all the other journeys that are still collectively very important?"
As Khachatryan points out, businesses may not be able to deploy a data engineer for every campaign. So the new decisioning product will aid marketers who don't have that luxury.
The scale on offer is considerable. A single agent can test more than a billion combinations simultaneously, learning and adjusting continuously rather than running a fixed A/B test that takes months to reach statistical significance.
To put that in perspective, testing hundreds of creative variations the traditional way would usually mean a marketing operations team producing each version by hand, then a data science team running sequential A/B tests to work out which one performs best, a lengthy process that still only identifies one winning version for everyone, rather than personalizing anything at the individual level.
Khachatryan pointed to beta customers who have already doubled campaign performance using the tool instead.
"Now you can spin up 500 creatives in like two minutes. And then have 500 and go immediately test them against each other. You're not just finding the best one, you're actually personalizing at the individual level"
Decisioning Studio Go is in beta now and expected to reach general availability on October 14.
Agentic Standards: Quality Assurance That Runs on Autopilot
Speed creates its own bottleneck. If AI lets a marketer build a campaign in minutes, a manual compliance review at the final checkpoint becomes the slowest part of the entire process.
Agentic Standards is Braze's answer: an AI-powered review layer that checks campaign setup, copy, links, personalization logic, and brand or regulatory compliance before anything reaches a customer, flagging issues with pass, warning, or fail status and automatically suggesting fixes.
Not only does this speed things up, but Khachatryan identified the employee experience angle:
"No marketer, I think, wakes up in the morning and says, I'm so excited to QA stuff. What are marketers excited to do? They're excited to think about marketing strategy, to come up with ways to talk about the product"
Sportsbet, one of the beta customers named in Braze's announcement, said the tool had already closed real operational gaps, catching missing unsubscribe links and eligibility segment errors before they went out. For a regulated business, that could be the difference between a caught mistake and a compliance incident.
Agentic Standards is in beta and expected to reach general availability in October.
Operator Connect: Bringing Braze Into the Tools Marketers Already Use
Braze's AI assistant, BrazeAI Operator, already lets marketers build and configure campaigns inside the platform through natural conversation rather than manual setup. Operator Connect extends that same intelligence into external tools marketers use every day, including Claude, ChatGPT, Snowflake Cortex Code, and Microsoft Copilot, letting teams draft, troubleshoot, and launch Braze campaigns without leaving those platforms.
