Build Powerful
Agentic AI Workflows
AI Agents for Work is a new, hands-on program empowering non-technical operators to identify, design, and build agentic AI workflows that supercharge their work. Founding cohort launches November 13, 2026.
%20(1).png)
Unlock the power of AI agents—without the hype
AI tools are getting more capable at an astounding pace. Increasingly, they can gather information, make decisions autonomously, use tools, and carry out multi-step work in a way that offers real leverage.
You don't need to become an AI engineer to access these capabilities. You need to understand what these systems can do, where they break down, and how to design workflows that give your agents the context and data they need to be useful.
Built to fit your week
.png)
Who This Course Is For
This course is designed for non-technical professionals who want to stay ahead and unlock the ability to build powerful agentic AI workflows that drive real impact.
%20(2).png)
A few examples of what people build, by role:
MARKETING AND CONTENT: competitor digests, campaign recaps, content repurposing
OPERATIONS AND STRATEGY: request triage, vendor research, weekly status reports
PRODUCT, DESIGN, AND RESEARCH: research synthesis, feedback tagging, competitive teardowns
CONSULTANTS: pre-meeting briefs, market scans, first-draft proposals
FOUNDERS AND ENTREPRENEURS: business snapshots, investor update drafts, inbox triage
Don't see your role? If your work involves gathering information, making sense of it, and acting on it across multiple steps, you're in the right place. You should be comfortable with tools like ChatGPT or Claude, but you don't need to know how to code.
Build an Agentic Workflow Around Your Real Work
This isn't a course where you just watch someone else build demos. Over four weeks, you'll identify a workflow from your own work and turn it into a functioning agentic system.
YOU WILL LEARN
A Course Designed Around Building
Understanding agents conceptually is useful. Building one is where things get messy.
That's why this course includes multiple layers of support as you move from idea to working workflow.
WEEKLY LIVE LECTURES
Join a 90-minute live lecture each week covering the concepts, frameworks, examples, and techniques you'll need for that stage of your project.
SMALL PEER GROUPS
Later in the week meet with a small group of classmates to pressure-test ideas, share progress, and learn from people applying the same principles to different types of work.
HANDS-ON PROJECT WORK
Rather than completing disconnected exercises, you'll progressively design and build one useful agentic workflow around your own work.
SUPPORT WHEN YOU GET STUCK
Participants wil have access to live office hours and async support when a specific issue is keeping you from moving forward.
The goal isn't to build the workflow for you. It's to help you get unstuck while developing the confidence to solve these problems yourself.
Enrollment opens soon.
Join waitlist and get a special discount.
Four weeks to your first working AI agent
Understand Agents & Find the Right Workflow
Learn the difference between agents, chatbots, and traditional automations — and understand where agentic systems are useful today, and where they still struggle. Identify several potential workflows from your own work.
Design the system
Gain an understanding of agentic workflow best practices, then select one of your candidate ideas into a clear workflow before you start building.
Connect & Build
Explore the tools necessary to connect your agent to your data and systems. From there, turn the workflow blueprint you created the previous week into a functioning system.
Test & Improve
Dive into testing, reliability, guardrails, and the right place for human oversight. Apply these frameworks to what you've built, identify where it fails, and improve it before putting it into real use.
Experience helping professionals level up with AI.
New fields.
Proven ways of learning.
Workbench comes from Designlab, where more than 25,000 learners have built real skills in UX, design, and product, and gone on to do that work at leading companies like Google, Dell, McKinsey, and many more.

AI is now changing every kind of work, not just design or product. Workbench is our effort to bring what we've learned about teaching practical AI skills to more professionals, wherever they sit.
The tools are new. The way people actually learn hasn't changed.
YOU LEARN BY DOING
Every program is built around hands-on project work, not passive videos or disconnected quizzes.
LEARN FROM PEOPLE DOING THE WORK
Our instructors and mentors are practicing experts, applying these tools in their own jobs every day.
NEVER LEARN ALONE
Live sessions, peer groups, and direct feedback keep you moving when you get stuck.
50+ Countries
Clients
& Instructors
Earn a Certificate of Completion
Complete the course and your project to earn a certificate recognizing your work in AI agents and agentic workflow design.
.webp)
Still deciding?
No. AI Agents for Work is designed specifically for non-technical professionals. You'll design and build a working agentic workflow using AI tools and no-code connectors. No coding is required.
You don't need to be an expert, but you should be comfortable using AI tools. Most students regularly use ChatGPT, Claude, or a similar assistant for everyday tasks like drafting, research, or analysis.
You don't need any experience setting up AI workflows or working with agents. That's what the course teaches.
An AI agent is an AI system that works toward a goal on your behalf. Rather than just responding to a single prompt, it can gather information, make decisions, use tools, and carry out multi-step tasks.
In this course, you'll learn how to design workflows that give agents the context, data, and human checkpoints they need to do useful work reliably.
AI is moving fast, and some specifics, like individual tools, features, and interfaces, will change. That's why this course focuses less on any single tool and more on durable skills: identifying which work can be delegated, designing workflows around a real goal, giving agents the right context, and evaluating whether the output is actually good.
Those frameworks transfer as the tools evolve, so what you learn should keep paying off long after the course ends.
You'll build at least one useful, multi-step workflow drawn from your own work: something you or your team does regularly, like researching and summarizing information, preparing reports, triaging requests, or moving information between tools.
For a first workflow, simpler is better. A strong project matters enough to be worth automating but is scoped tightly enough to finish in four weeks. It also needs to connect to systems and data you have access to. We'll work with you early in the course to choose a workflow that's both useful and realistic.
It happens, and we'll help you work through it. Between your peer group, office hours, and the course Slack, you'll have support diagnosing problems and finding a way forward.
Some constraints are real, though. If a workflow depends on data or systems you aren't authorized to access, or on integrations that aren't practical yet, we'll help you rescope it or choose a different workflow that's both useful and executable. Learning to make that call is part of the course.
This isn't a tool-specific course. We teach frameworks for designing agentic workflows that work across platforms.
That said, you'll build your workflow in a real tool. We expect most students to use Claude/ Claude Cowork, ChatGPT Work, or Microsoft Copilot as their core agent. Depending on your workflow, you'll also connect it to your data and apps, usually through the connectors built into those tools and sometimes through automation platforms like Zapier or Make.
We expect most students will need a paid plan for their core AI tool, such as Claude, ChatGPT, or Microsoft Copilot, and some workflows may require an additional connector or automation tool.
For most projects, these costs are modest. If your workflow runs at a larger scale, costs may be higher, and we'll help you plan for that when you scope your project.
AI Agents for Work is built and taught by a team. Chadwick Smith, who leads Designlab's internal learning experience team, is developing the curriculum in collaboration with external experts who have deep, hands-on experience building AI workflows.
Throughout the course, you'll learn from several of these practitioners, including the lecturer who leads the weekly live session and the expert facilitator who guides your peer group.
Plan for roughly 6 hours per week. That includes the 90-minute live lecture, a 60-minute peer group session, and async time working on your project. The exact amount will vary with the scope of your workflow.
On Mondays, we host a 90-minute live lecture. You'll be introduced to the concepts and techniques you need for that stage of your project and see live demonstrations. When you enroll, you'll automatically be registered for these sessions.
If you can't attend live, you can watch the recording instead. Just plan to watch it and complete the week's async project work before your peer group session. The same goes if you can't start on day one: you have some flexibility, as long as you're caught up before you meet with your peer group.
Peer group sessions are weekly 60-minute meetings with a small group of classmates, facilitated by an expert practitioner with hands-on experience building AI workflows. You'll pressure-test ideas, share progress, and learn from people applying the same principles to different kinds of work.
After you enroll, you'll receive an email to confirm your course details and choose your peer group time. Groups typically meet several days a week at 12pm ET (9am PT) and 6pm ET (3pm PT), and you'll meet on the same day and at the same time each week.
Beyond the weekly live lecture and your peer group sessions, you'll have access to office hours and a course-only Slack community where you can ask questions, get unstuck, and share progress with classmates and the course team.
Yes. Students who successfully complete AI Agents for Work, including the final project, will receive a Certificate of Completion.
We'll be offering a special discount to our Founding cohort when enrollent opens — but seats to this first cohort will be tightly capped.
We also offer a payment plan that lets you pay half up front and half at the end of the course. To learn more, contact our admissions team at advanced.admissions@designlab.com.
Absolutely. Many students have their tuition covered by their employer's learning & development budget. If you have questions about the process, email advanced.admissions@designlab.com.
To help you request reimbursement, feel free to copy, paste, and edit the email below:
Hi {Manager Name},
I'd like to use my learning & development budget to enroll in AI Agents for Work, a 4-week live course from Workbench, a division of Designlab, who has trained more than 25,000 learners over 13 years.
The course is designed for non-technical professionals and focuses on applying AI agents to real work. Over four weeks, I'll learn to:
-Identify work where AI agents can add real leverage
-Design an agentic workflow around a real goal, with the right context, data, and human checkpoints
-Build a working version using modern AI tools, no coding required
-Test and refine it so it actually improves my work
The course includes weekly live lectures, small peer group sessions led by expert practitioners, and hands-on project work. I'd build my project around a workflow from my own job, so the benefit to our team would be immediate.
The course costs $1,199 USD. You can see the full details here: [link]
Thanks,
{Your Name}
We’re happy to work with a team or company. Contact us at advanced.admissions@designlab.com for information on multi-seat discounts, custom trainings, and more.
AI Agents for Work is backed by the Designlab Happiness Promise. After you've fully participated in the course, if you're not satisfied, you can get your money back. More details on that here.
.png)
GET MORE DONE WITH AI AGENTS
Learn how to identify the right opportunities, design powerful agentic workflows, and build one around your own work.
