Ben McCulloch

Designers: vibe code a VUI prototype in 30 minutes

Conversation AI Designer

Vibe code a VUI with the same core capabilities as Amazon Alexa and Google Assistant in half an hour

Build a working voice assistant prototype rapidly

Don't be held back if you can't write code

Become a designer who builds with AI

Utilise AI code generation to become a designer who leads with working prototypes

Enhance your toolkit and employability

✅ Why invest in learning with me? I’m a Lead PM at Booking.com, and one of the first tech PMs there, with over 10 years in the field. Before moving into product, I was a Software Developer and earned a PhD in Computer Science. I’ve worked on both sides and know exactly what technical understanding PMs need - and don't. I'll teach you exactly what matters (and why) for driving products forward. You'll get the technical and AI-building fluency needed to drive your team, no coding, no database internals, no load balancers. ✅ What can better technical understanding do for your career? From a survey of 650 PMs, 56% want to speak the same language as their engineers to communicate effectively, and build trust. 61% want to estimate features on their own, so they can plan better and deliver on time. And today, technical understanding also lets you quickly AI-prototype features to show it to real users. Finally, if you want to land a Big Tech PM role, tech skills are critical. ✅ What makes this course different? ➡️ The course is 20% theory and 80% hands-on:&nbsp;you’ll solve over 200 exercises, because that’s how real skills stick. (Try two free lessons below to learn new things even if you don’t enroll) ➡️ The course naturally offers you two very related skills: Tech foundation (part 1) allows you to lead your team more effectively. Then, if you think about it - AI is... just a non-human version of your team :) That's why in part 2 you apply Tech concepts to create features yourself. This enables you to understand tech even deeper and later turn your ideas into clickable features in minutes/hours (check 2 examples below: Spotify Clone and Cost Split App). In student's words: "Interestingly, the AI part is also incredibly useful even if you’re not the one doing the prototyping yourself.&nbsp;I started using the right technical terms and began to think a bit like an algorithm (which helps in finding a common language with developers and engineers)." ➡️ You can join even if you have zero knowledge about tech or AI-building ➡️ Over 1,500 students have already completed the course online or in person, on this platform and others. See reviews below. ➡️ Even more valuable practice: all alumni get a 100%-off voucher for one of the other ProductDo simulators worth up to 400$. ✅ The course structure explained 1️⃣&nbsp;Live session: Tech 101sWe start with a Tech 101 live session.In just 2 hours, you’ll learn (and immediately apply) the core concepts every PM should know: Frontend, Backend, Services, APIs. If you only did this session, you’d already have the essentials down. 2️⃣ Practice Tech Core (50+ exercises)Time to roll up your sleeves. You’ll work through a story-driven series of exercises based on real-life scenarios. You’ll build a Taxi app and along the way you’ll master tech topics with your virtual colleagues: Daniel the bossy CEO, Jane the guiding Principal PM, Ivar the eager Jr. PM (who always needs help), Helen, your sharp Engineering Manager, and more. Curious if this hands-on style suits you?Grab the free resource below, which is one full lesson from the API chapter, and see if solving 200+ exercises like this one sounds fun. 3️⃣ Live session: building a product architecture togetherOnce you’ve mastered the theory and practiced solo, we'll meet up again.You’ll team up with other PMs to design an architecture, APIs, and more for a product idea. Students love this collaborative part! 4️⃣ Practice AI-prototyping (50+ exercises)Back to hands-on mode this time with an AI-prototyping simulator.- Start by cloning UI with simple prompts, then add more UI features.- Next, bring in the database — suddenly it starts to feel real.- Layer on backend logic to make it smarter.- Finally, integrate an external API like OpenAI so your app becomes even more powerful. 🤖 Two examples of what you can do with AI:- In 5 mins: the Spotify clone: now in another 10 mins I can imitate "Concerts nearby" feature and show to stakeholders instead of slides: https://productdo-spotify-clone-interface-demo.replit.app/- In 3 hours: Fully functional CostSplit full-stack app: https://cost-split-mobile-demo-no-auth.replit.app/ (this app is part of the course, you'll build it yourself). Want a taste? Grab another free resource below; it's the full lesson on UI feature creation. After this, you'll be able to add a feature to your Spotify clone all by yourself! 5️⃣ Live session: building a product from scratch using AINow we bring it all together. I’ll walk you through building a complete app, step by step. You’ll break into groups to handle the UI, backend, databases, APIs, and more. It’s going to be a lot of fun. 6️⃣ Bonus- You’ll unlock advanced simulator chapters, allowing you to continue progressing at your own pace.- We have one more workshop reserved for any other topics the group wants to dive into, plus an extended Q&amp;A.- Additionally, you’ll receive complimentary access to one of the other ProductDo simulators worth up to 400$. ✅ Prerequisites to join the course- No prior tech or AI experience needed- If you want to participate in the AI-prototyping part, you need a Replit account (25$ for a month should be more than enough)- Be ready to set aside 2 hours for the live workshops and 3-4 hours during the week for self-practice. Only through practice can you really learn; there is no way around it.

The technical understanding a successful PM needs - nothing more.

Course syllabus

Live workshops where we solve practical exercises together and in groups

Live workshops

Hands-on practice in the Simulators ("Tech for PMs" and then "Full-Stack AI Prototyping"), where you learn by doing

Simulator practice

Just send me a DM message in the Maven chat

Questions for the author

Course schedule

Non-tech Product/Program Managers who want to communicate effectively with their teams and assess the real complexity of projects

PMs, designers, and marketers to prototype features in minutes/hours or even start their own side project

Entrepreneurs/founders who want to build confidence in a tech-driven world, create an MVP on their own, and hire the right team later

Who is this course for

What you’ll get out of this course

Understand all core concepts of tech and lead the team with "what/why" without getting into "how"

Tech fluency to lead your team effectively

Build the ability to gauge development complexity, keeping your plans realistic and your launches on track

The skills to accurately estimate project scope and timelines

Get ready for Big Tech PM interview questions like “Can you design the architecture for a Tinder/WhatsApp/Uber/etc?”

Preparation to handle technical interviews

Clone your product’s UI, add interactive features, and wire up backend logic. All without writing code.

The ability to turn ideas into clickable prototypes

Use AI to create full-stack apps (UI, backend, database, APIs) in just hours or days. No engineering background or coding needed.

The power to build an MVP or start a side project on your own

Two related skills in one course: master tech to speak the same language with your team. Use it to AI-prototype features or a side project.

Oct'25

Vladimir Kalmykov | Lead PM @ Booking.com

Lead PM @ Booking.com, PhD, 1500+ PMs learned Tech with me

Cohort Aug'25

Jan'25

TL;DR:You don’t need a tech co-founder or coding skills to bring your idea to life.In just one week, you’ll learn how to validate, build, and test your startup MVP using the most powerful AI tools available today — no technical background required. There has never been a better time in history to become an entrepreneur.With the right guidance and AI-powered workflow, you can turn your idea into a real, working product faster than ever before. You’ll get direct, personalized guidance from Elena Avramenko - a seasoned founder, startup advisor, and hands-on product &amp; marketing generalist who has helped 100+ entrepreneurs validate ideas, launch products, and get early traction. This fast-paced bootcamp gives you practical experience across the full startup cycle — from discovery and validation to MVP building and go-to-market testing — all powered by modern AI and no-code tools. You’ll walk away knowing which tools to use (and which to ignore), how to build a functional MVP, design a simple user journey, and test with real users — all without writing a single line of code. Accelerate your path to a working, testable prototype and save 3–6 months, equity, and money. 💡 Today’s reality for founders:Don’t let your idea stay on the shelf — this is the best moment to make it real. ⚡ What You’ll Do✅ Shape and validate your ideaDefine your problem, audience, and value proposition before building anything. You’ll use AI-powered frameworks and a custom ChatGPT workspace to explore every angle of your idea.✅ Build your MVP with modern no-code AI toolsBring your idea to life using Lovable, and supporting tools like Supabase, Tactiq.io and Notion AI — fast, visual, and collaborative.✅ Use AI for every step of creationMaster how to apply AI for ideation, research, prototyping, UX writing, and storytelling.✅ Test and learn like a real product teamBuild your MVP and get real user feedback, and refine your product with insights.✅ Understand what makes a startup-ready product In clear, simple terms — from user problem to first traction. 🧭 What Makes This Course DifferentYou’re not watching tutorials — you’re building with expert guidance.Get live, hands-on support from Elena and join guest sessions on Go-to-Market, Design as a Differentiator, and Positioning Fundamentals. Elena is not just an AI tools expert — she’s a founder who has guided 100+ entrepreneurs from idea to first traction and serves as an Expert Mentor in Techleap.nl, the Dutch scale-up community of 300 founders.You’ll leave with tested tools, real skills, and a working product — not just inspiration. ✨ Outcomes You Can ExpectA functional MVP you built yourself that real users can try (and even pay for)A validated value proposition backed by AI-powered customer insightsA launch-ready marketing page to test with users, investors, or devsConfidence and clarity in using AI tools to think, test, and build like a founder 🧩 Real Examples Built in a Week by Non-Technical Founders- memelon.kids- kidswhovibecode.com- puurfect.lovable.app Each started as a simple idea — and became a live product in days.

The startup you always wanted to build? Time to make it real.

Get help in community of builders and course organizers

Anytime async

Feel free to type out dates as your title as a way to communicate information about specific live sessions or other events.

May 7, 2022

Schedule items can also be used to convey commitments outside of specific time slots (like weekly projects or daily office hours).

Weekly projects

Idea-stage solo founders who have no technical background but want to finally bring their vision to life.

Professionals exploring side projects who want to validate an idea before going all-in or quitting their job.

Curious builders and creators who love experimenting with the newest AI and no-code tools to build something useful — or just for fun.

Turn your idea into a working prototype using AI and no-code tools — ready to test or pitch.

Build your MVP

Learn how to define your audience, problem, and value before building anything.

Gain product clarity

Use ChatGPT, Lovable, Notion AI, and more for ideation, prototyping, and storytelling.

Master modern tools

Get proven frameworks for validation, user research, and go-to-market basics.

Learn the startup playbook

Stay connected with other founders for feedback, accountability, and next steps.

Join a builder community

Turn your idea into a working prototype in 7 days using AI & no-code tools. Learn, build, and launch—no tech skills required.

Miro

LEGO®

Readymag

Microsoft

Elena Avramenko

Shipped 100+ projects| Bootstrapped to 1M+ downloads| Product&Growth generalist

The first cohort and fastest path to success

But most teams are still designing like it’s 2015. AI is transforming industries and changing the very foundation of how products work. Companies are moving fast to integrate large language models, agents, and adaptive systems into their products. But as model complexity grows, the user experience often gets left behind. Powerful algorithms can now predict, generate, and decide but without thoughtful design, they often confuse users, introduce risk, or fail to earn trust. The challenge? Designing for AI isn’t like designing for traditional software. It’s not just about usability or polish anymore. It’s about system behavior, adaptation, and collaboration between humans and machines. If you’ve ever launched an AI feature only to see users abandon it because they didn’t understand what it was doing or if you’ve been asked to “design the AI experience” but weren’t sure where to begin → this course is for you. The Role of the Designer Is ChangingUntil recently, designers focused on interfaces and flows. Today, the frontier is different. OpenAI recently posted a job for a Model Designer, a new kind of design role focused on shaping how models behave. And it signals a broader shift across the industry: Designers aren’t just styling outputs anymore. They’re architecting how intelligent systems learn, adapt, and interact. In AI-powered products, design happens upstream, at the level of:- Data inputs and how users shape them- Model decisions and how they’re surfaced or corrected- Feedback loops that guide learning- System behavior that evolves over time To adapt to this new environment, designers must understand how models work, not to become ML engineers, but to ensure what the system learns, shows, or hides aligns with human expectations and values. Real Challenges This Course Prepares You For You launch an AI-powered feature and maybe it recommends content, summarizes input, or automates a task. But users are confused or drop off quickly. They can’t tell why the AI did what it did, or how to influence it. The model works, technically. The experience doesn’t.→ Learn to design transparency, feedback, and control into intelligent systems before trust is broken. Your CTO asks you to “design the AI experience.” There’s no interface yet, just a prototype from the ML team and a prompt to “make it usable.” You’re left staring at system outputs, unclear where to start, or how much of the system behavior is even within your control.→ Gain the tools and frameworks to work upstream, shaping behavior, not just UI. You’re in a meeting where the team is debating whether to launch a model with 72% accuracy. Product wants speed. ML is confident the model will improve post-launch. But no one’s talking about the UX cost of being wrong nearly 1 in 3 times or how the system will recover from those mistakes.→ Learn how to design for failure, uncertainty, and system correction from day one. A personalization feature goes live. But user feedback is harsh: “This feels random.” You realize the issue wasn’t the model, it was the lack of communication and guidance. There’s no explanation, no context, and no way for users to improve the output.→ Understand how to design model feedback loops and trust-building interactions. You inherit a half-built AI assistant. It occasionally interrupts users at the wrong time, answers inconsistently, and disappears when needed most. The UI looks fine but the behavior feels off. You realize this isn’t a layout problem, it’s a systems design problem.→ Move from UI design to shaping system behavior, flows, and agent logic. You will bring your own real-life project scenarios, and we’ll coach you through how to solve them—using frameworks, feedback, and hands-on critiques.

The AI Tech Revolution Is Redefining Design

Class will meet once a week for live lectures and project discussions.

Weekly Live Sessions

Plan to work on a capstone project to apply your learning to a capstone project

Weekly assignments

Get weekly 1-1 office hours for personalized feedback on your capstone project

1-1 Coaching 

Senior UX &amp; Product Designers: Deepen your systems thinking skills and learn to design around model behavior

Design Leads &amp; Heads of UX: Guide teams, shape AI product strategy, and lead with system-level thinking and trust design.

AI-Aware PMs &amp; Researchers: Collaborate better with design, align model goals with UX, and support trust focused, adaptive systems.

Learn how to speak the language of models, data pipelines, and evaluation metrics—so you can contribute meaningfully from early scoping to final delivery. You’ll practice framing use cases, defining handoffs, and aligning design with system capabilities.

Collaborate effectively with ML and Product teams

Instead of just designing around the model, you’ll explore how interaction design can shape what the model learns, retains, or ignores. You’ll build UX patterns that create useful training signals, structured inputs, and adaptive guidance.

Shape model behavior through smart UX architecture

You’ll create flows that help users understand, question, or override the model’s decisions—like confidence indicators, correction mechanisms, and recovery paths for failure scenarios.

Design interactions that support user trust, control, and adaptation

Learn to design feedback loops where user actions (like corrections, inputs, skips, or feedback) help the model adapt. You’ll map those flows and create testable patterns that help close the loop between model performance and user value.

Build systems that improve over time—with users in the loop

Whether you’re dealing with ranking systems, generation models, or multi-step agents—you’ll practice mapping decision logic, designing trust dashboards, and identifying failure points. You’ll learn how to advocate for UX in model evaluation and behavior design conversations.

Influence how decisions are made, surfaced, and corrected at the model level

Go beyond interfaces—shape how intelligent systems behave.

Stanford University

Amazon

Google

Uber

Rupa Chaturvedi

Cohort 1

Cohort 2

We built this course after working with over five hundred students since August, 2023, 350+ of whom we put through the paces of our intense, 10-week flagship course,&nbsp;The AI Engineering Bootcamp: https://maven.com/aimakerspace/ai-eng-bootcamp Only 175 received a certification: https://bit.ly/aie-cert-process Why? Because even though students could complete&nbsp;The AI Engineering Challenge, that didn't mean that they&nbsp;really understood&nbsp;the core concepts and code required to build 🏗️, ship 🚢, and share 🚀 end-to-end LLM applications. You know, the important, boring bits. - How do you actually construct a&nbsp;front end&nbsp;using tools from the React Ecosystem (e.g., Next.js, Remix, etc.)? - What about a&nbsp;back end&nbsp;using Python (e.g., FastAPI, Flask, Django, etc.)? - What about&nbsp;connecting them&nbsp;using AI-assist with industry-standard version control and branching? -- This was true before vibe-coding, when we used to use Chainlit for front ends and Hugging Face Spaces to deploy back ends. These days, it's all about vibes. It's all about AI-assist and using coding agents like Claude Code, Codex, Cursor, Devin, Manus, and the like. But AI-assist doesn't help you understand how to fix things when they go wrong. In fact, in a world of AI-assisted development, the iteration process itself includes vibe coding, especially early on in the development cycle. But alas, sofware engineering is still software engineering, even if it's building AI applications, building with AI assist, or both! --- e.g., When things go bad, you still need to be able to revert back. When you need to add a feature to upgrade your app - like PDF upload for RAG or basic Agentic RAG - you need to be able to do this without ruining your existing code. Vibe coding gets people into trouble AI doesn't hold context like we do as humans. This has many implications, from the second wave of MCP to the future of product management upstream of the actual engineering, but those are all advanced topics we cover in our 10-week bootcamp. -- This course is about giving you the time and space to get the fundamentals right; and to learn the best practices for coming up to speed in an AI-assisted world, insofar as they exist already. There is a great leveling occuring; that is, the people who are sophisticated software engineers are picking up AI-assisted tools like coding agents off the shelf and using them often just as much as beginners. In 2025 heading into 2026, the vast majority of developers and future developers are all starting from a similar spot, and there has never been more demand for these skills. This course will make sure you understand what that spot is, and how you can make the most out of the opportunity space for your career as the AI-powered future ahead continues its steady drumbeat of progress into the new year. Welcome to the The AI Engineer Onramp.

Build LLM apps with coding agents like a professional AI Engineer

Mondays

Each class period, we will get hands-on with Python coding homework!

Weekly Programming Projects

We typically have 10+ hours per week of office hours that you are able to attend (1 for each instructor and peer supporter!)

Instructor & Peer Support Office Hours

Aspiring AI Engineers who do not yet code every day, but want to code every day.

Aspiring AI Engineering Leaders who want skin in the game on building, evaluating, and deploying at least one end-to-end LLM application.

Potential Aspiring AI Engineers&nbsp;who do not yet code every day, but think they might&nbsp;want to code every day.

Who this course for

Set up your workstation like a professional AI Engineer, including your own specific rules, docs, package specifications, and git workflows.

🧑‍💻 Set up a professional AI-Assisted Interactive Development Environment (AI-IDE)

Learn where the vibes go bad when vibe coding front ends, back ends, connecting them, and making modifications!

🎧 Learn best (and worst) practices for Vibe Coding with Coding Agents

Learn to debug AI-generated code effectively.

🕳️ Learn How to avoid vibe-coding yourself into a hole

Understand how front-end, back-end, and LLM components interact

🚢 Build and deploy three end-to-end LLM applications

Develop your own style of vibe checking, preparing you to learn more systematic evaluation techniques in the future.

📊 Vibe-check and do cursory evals of your LLM apps

Master Git workflows using best-practice branch development, and start seeing those commits!

🐙 Finally set up your personal GitHub 

Be ready to tackle our 10-week flagship course <a href="https://maven.com/aimakerspace/ai-eng-bootcamp" rel="noopener noreferrer" target="_blank">The AI Engineering Bootcamp</a> (or any other!) with confidence.&nbsp;

🥇 Learn the prerequisite skills needed to become an AI Engineer in '26

Build and ship end-to-end LLM applications using established and emerging AI-assisted development best practices for coding agents.