Free 5-Day Live Cohort • 50 Seats Only
Opal U is a live 5-day cohort where 50 senior experimentation practitioners build 3 AI agents that save 10+ hours a week and remove bottlenecks slowing their programs down.
Join the AI experimentation cohort to build and ship agents for your experience optimization workflow.
300 Opal University graduates to date
3,000 professionals trained
20% promoted or landed new roles
1,667 agents now in the Opal directory
Trusted by 9,000+ business and millions of happy customers
It’s been an incredibly rewarding journey — weeks of hands-on learning, from prompt engineering to building and deploying production-ready AI agents. 🚀 A huge thank you to Optimizely for delivering such an outstanding programme. I’m especially proud to be part of the very first cohort! 💙 Special thanks for making the entire learning experience so engaging, supportive, and inspiring. 🙌 Feeling proud, thankful, and very excited for what lies ahead — the future with AI looks incredibly bright. ✨
CRO Strategist
I'm Stella. I'm leading Opal U, Optimizely's AI education platform. I help experimentation leaders go from AI-curious to AI-confident, building agents that actually change how they work, and where their careers go next.
I'm Giulia. I've spent the last two years building with AI in public. I've built AI content for teams at Inditex, IBM, Zapier and Perplexity, from content intelligence pipelines to executive AI fluency programmes.
Our graduates include experimenters from companies like Unicef, BBC, Mattel, HSBC, Accenture - everyone from CROs, experimentation managers to VPs.
Wherever you're starting, there's a seat here for you.
No theory. No fluff. Just 3 working agents that save you real time starting Monday morning.
Save 3-5 hours/week
Your first agent that automates a repetitive task from your experimentation workflow. Most practitioners save 3 to 5 hours a week with this one alone.
Example: Summarize experiment outcomes, surface patterns across historical results without waiting for an analyst to pull it together.
Your "holy s&*!" moment
A more complex workflow that connects multiple steps in the testing cycle. This is where the "why haven't we always done it this way" moment happens.
Example: Map hypothesis → Prioritize backlog → Generate variants → QA experiment → Analyze results → Recommend next test
Fix your biggest bottleneck
Your custom agent tailored to YOUR biggest bottleneck. This is what you'll demo on Friday and share with your team.
Example: Build a "results storytelling pack" agent that turns experiment outcomes into an exec-ready results pack.
1 hour per day. Live training. Build together.
Map your biggest time sucks. Learn the framework for deciding what to automate first. Walk out with 3 agents you'll build this week (no analysis paralysis).
Stop writing prompts like everyone else. Learn the 5 techniques that turn "meh" outputs into "how did you do that?" results. Build Agent #1.
Your agents need to know YOUR business, YOUR brand, YOUR audience. Learn how to feed context so your outputs actually sound like you. Build Agent #2.
Connect your agents to real data sources. Set up quality checks so you're not babysitting every output. This is where it gets powerful. Build Agent #3.
Show off what you built. Get feedback. See what others created. Leave with 10+ new ideas to build next. (And a LinkedIn post that gets you noticed.)
This cohort is designed for senior experimentation leaders who need results fast.
A Senior, VP, or Head of Experimentation/ CRO/ Digital who needs to prove ROI from AI
Overwhelmed by AI hype and don't know how to scale the program without scaling the team
Expected to "do more tests with less" and need to multiply your output
Ready to dedicate 1 hour/day for 5 days to transform how you work
Prepared to become the AI expert on your team who knows how to build and ship agents
You're just exploring AI casually (this is hands-on building, not theory)
You can't commit to live sessions (recordings available, but you'll miss the magic)
You're looking for a certification to put on LinkedIn (we give you working agents, not badges)
You want to learn "AI strategy" without building anything
Join the next cohort. Walk out with 3 agents that save you 10+ hours every week!
Only 50 seats
