In this episode, we chat with Sid from Composio to understand what MCP (Model Context Protocol) really means and how it's changing the way AI tools connect with different apps. Sid explains how MCP helps simplify complex processes and makes it easier for developers and AI agents to access the tools they need.

We also talk about how this technology works behind the scenes, how it compares to APIs, and how big companies like Microsoft and Anthropic are getting involved. Sid shares real examples from Composio, showing how context-aware tools are being built and why they matter.

Whether you're a developer, someone working with AI, or just curious about where this technology is headed, this conversation offers clear explanations and useful insights into a rapidly growing space.

0:00 – Intro & Sid’s Background

1:30 – What Is MCP (Model Context Protocol)?

3:50 – Is MCP a tool

5:40 – What MCP Includes (Resources, Tools, Prompts)

8:55 – Journey of MCP

15:33 – Example of MCP In action

25:00 – Developer Experience & Setup

47:00 – The Future of MCP

Jacob Bank, CEO of Relay and former product lead at Gmail, joins Aishwarya Ashok to explore how AI agents are truly reshaping the way we work. He explains why structured AI workflows solve most real-world needs today and highlights the concept of “spiky intelligence,” where AI excels at some tasks but still needs human direction. Jacob emphasizes that the most important skill in this new era isn’t coding, but the ability to clearly define goals and structure problems for AI. He shares practical ways to use AI not just to replace tasks, but to automate complex, cross-platform outcomes—unlocking new levels of productivity and enabling small teams to achieve big results. 00:00 - Intro 1:17 - Welcome 2:34 - Jacob's Journey: From Google to AI Agents 3:25 - Beyond Tools: Focusing on Workflows & Outcomes11:34 - AI Workflows vs AI Agents: What's the Difference?19:43 - The Human Role: Critical Thinking & Future Skills27:30 - Understanding "Spiky Intelligence" in AI31:30 - How to Start Building Your Own AI Agents43:21 - Demo of Relay53:11 - Summary and Outcome of the agent

Ami Vora, CPO at Faire and former VP of Product & Design at WhatsApp, shares a unique perspective on how fear can be a driving force in decision-making and career growth. Raised in a family where fear played a significant role in shaping decisions, she turned fear into a tool for clarity.Rather than ignoring fear, she acknowledges it, writing down her fears and actively mitigating them when making big decisions, such as accepting a new job. By making the fear explicit, she unravels its power, turning it into actionable insights that help guide her.

Vora's experience reflects the ever-changing landscape of tech, where continuous learning is the only way to remain relevant. She draws on a lesson from a high school teacher, emphasizing the importance of staying on the frontier of knowledge to adapt to the unpredictable nature of the industry. Vora integrates this mindset by constantly challenging herself with low-stakes learning experiences outside of work, like ice skating, to reinforce the mindset of growth and adaptation. She reminds us that, especially in product and AI, there’s a need to embrace curiosity and let go of the pressure to always be right. In a world where everything is evolving rapidly, embracing failure and being comfortable with not knowing is essential for continual progress.Chapters0:00 Technology Won't Save You—Learning Will

2:09 Meet Ami Vora: From WhatsApp to Wild Versatility

4:04 How Jokes Helped Me Lead Through Hard Times

6:18 Turning Liberal Thinking Into a Career Superpower

9:00 What Actually Drives Us? Energy Over Everything

13:00 The Hardest Job I Stayed In—and What It Taught Me

17:00 Fear, Feedback, and Showing Up Anyway

20:00 Context Switching for Humans (Not Just AI)

22:08 The Tools I Trust: Reframing, Fear-Setting, and Closing Loops

36:02 AI, Agency, and Always Being Behind

43:27 Being Right vs Being Curious—The Career Mindset Shift

47:13 Execution > Strategy: Why Doing Beats Planning

51:25 My Career Is a Small Plates Dinner—and That’s a Superpower

Responsible AI is no longer just a theory—it's a necessity. In this episode of The Founder's Foyer, we sit down with Aishwarya Srinivasan to explore how organisations are tackling ethical AI in real-world environments.

This discussion covers deep and practical insights into data privacy, copyright, enterprise integrations, and the future of agent-based systems. If you're building with AI or implementing LLMs in your org, this episode is a must-watch.

Learn how AI adoption, compliance, and governance must go hand-in-hand to shape innovation responsibly.

Timestamps:00:00 – Introduction: Why Responsible AI Matters 02:42 – Data-Centric AI vs. Theory: Applying Ethics in Practice 07:35 – Copyright & Model Training Dilemmas 11:40 – Post-Training, Filters & What Data Should Be Used 16:10 – Role-Based Access & AI in Enterprise Workflows 22:25 – Productivity Tools & Data Privacy Risks 29:20 – Agent vs Automation: What’s the Difference? 35:10 – Human-in-the-Loop: Where It Matters Most 41:00 – Real AI Use Cases: Sales, GTM, and Non-Tech Teams 48:00 – The Need for Cross-Functional AI Literacy 52:45 – The Challenge of Data Silos and Integration 58:10 – Final Thoughts: Building Trustworthy AI Systems

How AI-driven development tools like Bolt are revolutionising the way entrepreneurs build and launch apps.

This discussion highlights groundbreaking real-world examples, from reducing costs by 99% to accelerating product development from months to weeks.

Learn how non-developers are creating full-stack apps and CRMs with ease, leveraging AI to bring ideas to life faster than ever before.

Speed is key, and Bold is leading the way in this new era of rapid innovation.

00:00 – Introduction

02:45 – The Viral Hooks App Case Study

06:10 – From $5,000 to $50: Cost & Time Revolution with Bolt

08:50 – AI-Powered CRM Case Study (Built in 3 Weeks)

15:40 – Speed is the Biggest Competitive Advantage

19:10 – Bolt vs Big Tech

23:00 – Real-Time Feedback & Prototyping with AI

26:00 – Bolt’s Open Source Edge & Model Integration

30:50 – End Users Don’t Care About Models—Only Results

33:20 – Bolt’s Rise to Market Leader in Web-Based Code Gen

36:15 – Bolt for Everyone: From Personal Sites to Full SaaS

39:00 – The Journey: Building Bolt & Going Viral

42:00 – What Makes Great Founders?

44:50 – The Vision: Building Apps as Easily as Designing in Canva

In this episode, Colin dives into the world of AI-powered coding tools and shares

actionable strategies for building projects efficiently. Whether you’re a developer or

a non-technical founder, learn how to choose the right tools (like Replit, Bolt,

Lovable, and V0), structure your workflow, and avoid common pitfalls.

*Why planning beats jumping straight into code

* When to use mock data vs. backend integrations

* Tool comparisons: Replit (full-stack), Bolt (prototyping), Lovable (non-technical),

and V0 (styling)

* The importance of validating ideas before scaling

Timestamp:

00:00 – Introduction

05:34 – Choosing the Right AI Coding Tool

09:47 – Why Planning Before Coding Matters

20:07 – Strengths & Weaknesses of AI Coding Tools

30:51 – AI for Non-Technical Founders

36:50 – Frontend vs. Backend: What to Build First

47:29 – The Reality of Building SaaS & AI Products

50:17 – The Smartest Way to Validate Your Idea

56:30 – The Future of AI Coding & Development

01:00:00 – Conclusion & Key Takeaways

00:00 - Intro: The Role of AI as a Thought Partner02:54 - Moving Beyond AI for Drafting: A Productivity Shift06:08 - Using AI for Conversations and Decision-Making09:48 - Project Knowledge: Uploading Context for Better AI Responses13:41 - The Truth About Prompt Engineering16:03 - Iterative AI Interactions: Refining Responses21:01 - AI as a Thought Partner: A Self-Driving Car Analogy29:46 - AI for Web Apps: Where Are We Now?37:27 - AI as a Nutritionist: A Surprising Use Case44:57 - Consuming News with AI: The Second Mouse Gets the Cheese

Tal Raviv has been a hands-on product manager & technical founder for 15+ years at hyper growth companies, always among the first PMs. He’s a Gen AI PM, and has been on the early teams of Patreon, Riverside, Wix, AppsFlyer, DuckDuckGo. He teaches one of the most popular PM courses in Maven focused on building your own AI copilot, which we’re going to dive into today.

We dive deep into how AI can be used as a thought partner, not just a tool for drafting documents. Whether you're a product manager, content creator, or AI enthusiast, this video will help you rethink how you interact with AI to maximize productivity, creativity, and decision-making.🔹 Learn why AI is more than just automation🔹 Discover the power of iterative AI interactions🔹 Understand how project knowledge enhances AI responses🔹 Explore unexpected AI use cases, from nutrition to filmmaking🔹 Find out why prompt engineering is overrated

00:00 Intro to Harold

01:30 Harold's Journey into No Code

04:11 From No Code to AI

06:34 Overcoming Analysis Paralysis in AI Building

10:43 Starting Small with AI

16:14 Building with AI Insights

25:57 Learning from AI Errors

27:28 The Importance of Practice in AI Learning

29:08 Building and Collaborating with AI

30:14 Personalizing AI Workflows

32:28 Challenges and Future of AI Tools

41:09 Advice for AI Tool Founders

If you're looking to get started with AI building, here's a special discount for you:

https://maven.com/no-code-ai/no-code-ai-bootcamp?promoCode=FOUNDERSFOYER

https://maven.com/no-code-ai/coding-with-ai?promoCode=FOUNDERS-FOYER

Product building has been democratized with AI—from PMs, developers and designers to exec leaders, everyone’s been eyeing on how to ride the AI wave and ensure they’re up to date with what it takes to grow in this new economy. The latest in that journey is how AI coding agents have been remarkable in helping people go from ideas to working prototypes in hours and days.

Harold Dijkstra is currently building the 100 School—making tech and AI skills accessible for anyone with ideas. With helping more than 41k+ builders across the globe, their online approach has been super helpful for creators and builders to ship prompts to products.