In this episode, we're joined by David Soria Parra, Engineering Lead at Anthropic and one of the core maintainers of the Model Context Protocol (MCP), to explore the biggest evolution of the protocol since its launch and why MCP is becoming the foundation for the next generation of AI agents.


What an Anthropic Engineer Thinks About MCP

Rosie is iFood's AI support agent, serving millions of customers across Brazil. Rather than trying to eliminate latency, the team designed the system around it. In this conversation, César Gomes explains how iFood uses asynchronous workflows, specialized agents, rollback strategies, delayed memory consolidation, and a reviser agent to keep conversations responsive while reducing costly hallucinations. The result is a production-ready architecture focused on one outcome: shortening the path from customer problem to resolution.

How iFood Designed Rosie Around Latency

Manish Dasaur is a Managing Director at PwC with over 20 years in data and AI, having helped 100+ clients navigate AI disruption and extract real business value from data, AI, and agentic AI initiatives. In this episode, he breaks down why most enterprise AI programs stall — and the playbook the winners are using instead.

AI Hype vs. Real Value

Loop engineering is the shift from manually prompting AI agents turn-by-turn to designing autonomous systems that do the prompting for you. Instead of writing a prompt, reading output, then writing the next prompt, you build a small control system that discovers work, delegates tasks to agents (and sub-agents), verifies results, persists state, and decides what to do next on a schedule or until a goal is met. It sits one level above agent harness engineering: where a harness equips a single agent run, a loop keeps firing agents on a cadence, spawning helpers, and feeding itself. The framing is increasingly tool-agnostic, as both Claude Code and OpenAI Codex have converged on similar primitives. The people building at the frontier are already living it: Boris Cherny, head of Claude Code at Anthropic, has said his job is no longer writing prompts but writing the loops that do.

A Discussion on Loop Engineering - AAIF Reading Group July 2026

The Model Context Protocol (MCP) is quickly becoming the standard for connecting AI assistants and agents to the tools, data, and systems they need to get real work done. As adoption accelerates across the open source ecosystem, developers are discovering new patterns, frameworks, and production practices for building MCP-powered applications.Join us for MCP Jam, a virtual community event bringing together developers, maintainers, and practitioners to explore the latest innovations around MCP. Through technical talks, live demos, and real-world case studies, speakers will share how they're building MCP servers, creating powerful tool ecosystems, improving reliability, enabling agentic workflows, and pushing the protocol forward.Whether you're just getting started with MCP or already building production systems, you'll leave with practical insights, implementation ideas, and a better understanding of where the ecosystem is headed.<h3>What to Expect</h3><ol><li data-list="bullet">Deep dives into building and deploying MCP servers</li><li data-list="bullet">Lessons learned from real-world MCP implementations</li><li data-list="bullet">Live demonstrations of MCP-powered applications</li><li data-list="bullet">Best practices for tooling, security, observability, and developer experience</li><li data-list="bullet">Community discussions with engineers shaping the MCP ecosystem</li></ol>This event is designed for AI engineers, software developers, open source contributors, platform teams, and anyone interested in the future of interoperable AI systems.Interested in speaking? We have a limited number of community lightning talk slots available. If you're building something with MCP, <a href="https://aaif.live/events/mcp-jam/talks" rel="noopener noreferrer" target="_blank">we'd love to hear from you</a>!

A virtual MCP Community Event for AI Builders

Join us for Session 20 of the Coding Agents Lunch &amp; Learn series as we continue exploring the rapidly evolving world of coding agents and the Model Context Protocol (MCP). We'll begin the session with a special featured presentation from Dimitrios Skarlatos, Assistant Professor of Computer Science at Carnegie Mellon University and founder of Lithos. Dimitrios will explore why inference speed is becoming the defining bottleneck for AI agents and demonstrate how serving frontier open-source models at over 1,000 tokens per second fundamentally changes what's possible for agentic workflows. Through a live demo, we'll see how high-performance inference can unlock faster feedback loops, more capable agents, and entirely new development patterns. Following the featured talk, we'll continue our exploration of the Model Context Protocol (MCP) and the rapidly growing ecosystem of MCP servers powering modern AI applications. After introducing the foundations of MCP in our previous session, we'll shift our focus toward practical applications. Together, we'll explore some of the community's favorite MCP servers, examine how they're being used in real-world agent workflows, and discuss what makes an MCP server useful, reliable, and production-ready. We'll also look at emerging benchmarks and evaluation efforts that help developers better understand the expanding MCP ecosystem. The second half of the session will be an open, community-driven discussion where attendees can share experiences, discover new MCP servers, and exchange best practices for integrating tools into coding agents and AI applications. As the ecosystem continues to evolve, understanding the available servers and how to evaluate them is becoming just as important as understanding the protocol itself. Whether you're building coding agents, AI assistants, developer tools, or simply keeping up with the latest advancements in agent engineering, this session is a great opportunity to learn from leading practitioners, ask questions, and exchange ideas with the community. Bring your favorite MCP server recommendations, your experiences building with MCP, and your questions, we're looking forward to another great discussion.

Coding Agents Lunch & Learn – Session 20: MCP in Practice

<h1>Loop engineering is the shift from manually prompting AI agents turn-by-turn to designing autonomous systems that do the prompting for you. Instead of writing a prompt, reading output, then writing the next prompt, you build a small control system that discovers work, delegates tasks to agents (and sub-agents), verifies results, persists state, and decides what to do next on a schedule or until a goal is met. It sits one level above agent harness engineering: where a harness equips a single agent run, a loop keeps firing agents on a cadence, spawning helpers, and feeding itself. The framing is increasingly tool-agnostic, as both Claude Code and OpenAI Codex have converged on similar primitives. The people building at the frontier are already living it: Boris Cherny, head of Claude Code at Anthropic, has said his job is no longer writing prompts but writing the loops that do.</h1> <h1>This month, David DeStefano, Sam Christensen, and Valdimar Eggertsson will sit down for an hour and chat about some of their learnings and experiences exploring loop engineering. We'll also open it up to the audience, who may have their own experiences and insights we'd love to hear more from.</h1> <h2>📖 Reading Material</h2><h2>How to Create Loops with Claude</h2><h2>Please read before the discussion (recommended):</h2><h1>https://todatabeyond.substack.com/p/how-to-create-loops-with-claude-a?r=1sqbmi&amp;utm_campaign=post-expanded-share&amp;utm_medium=web&amp;triedRedirect=true</h1>

A Discussion on Loop Engineering

Join us for Session 19 of the Coding Agents Lunch &amp; Learn series as we dive into one of the most important emerging standards in the agent ecosystem: the Model Context Protocol (MCP). As AI agents become increasingly capable, the ability to securely and reliably connect them with tools, services, and external data has become a critical challenge. MCP aims to provide a standardized way for models to interact with the systems around them, making agent development more interoperable, portable, and scalable.In this session, we'll explore what MCP is, why it has gained so much momentum across the AI community, and how developers can begin incorporating it into their own agent architectures. We'll discuss the problems MCP is designed to solve, examine real-world implementation patterns, and hear directly from the community about lessons learned while building and deploying MCP-enabled applications. If available, we'll also be joined by a maintainer from the MCP ecosystem to share insights into the latest developments and answer questions from attendees.As always, this Lunch &amp; Learn is designed to be interactive and community-driven. Whether you're building coding agents, AI assistants, developer tools, or simply trying to stay current with the rapidly evolving agent landscape, this session offers an opportunity to learn from practitioners, ask questions, and exchange ideas with others exploring the future of agent engineering.Bring your questions, your experiences, and your curiosity—we're looking forward to another great discussion.

Coding Agents Lunch & Learn – Session 19 - Understanding MCP

Session 18 is dedicated to one of the most valuable aspects of the Coding Agents Lunch &amp; Learn series: learning directly from the community. This special Community Show &amp; Tell session provides an opportunity for attendees to share the projects, workflows, tools, experiments, and discoveries they've been working on while exploring the world of coding agents and agentic engineering. The pace of innovation in this space continues to accelerate, and many of the most interesting breakthroughs come from practitioners experimenting with new ideas, combining tools in unexpected ways, and pushing existing workflows beyond their intended use cases. During this session, participants are encouraged to showcase what they've built, discuss challenges they've encountered, demonstrate useful techniques, and exchange lessons learned with fellow builders. Whether you're working on coding agents, multi-agent systems, evaluation frameworks, developer tooling, infrastructure, open-source projects, or entirely new approaches to software development, this is a chance to share your experiences and learn from others doing the same. No formal presentation is required—community discussion, live demonstrations, workflow walkthroughs, and informal knowledge sharing are all welcome. If you're looking for inspiration, feedback, or simply a chance to connect with other practitioners exploring the future of software development, this session is for you. Bring your projects, your questions, and your ideas as we close out another chapter of the Coding Agents Lunch &amp; Learn series together.

Coding Agents Lunch & Learn Session 18: Community Show & Tell

The Agentic AI Foundation (AAIF) is a Linux Foundation-hosted, vendor-neutral community creating open standards for transparent, interoperable AI agents across ecosystems.

Agentic AI Foundation

Home

Forum

Events

Content

Collections

People

Contributors

Albums

Clubs

Chat

Channels

Tools

Help

The Agentic AI Foundation (AAIF) is a Linux Foundation hosted, vendor-neutral community creating open standards for transparent, interoperable AI agents across ecosystems.

Your Privacy Choices

It looks like there is no account associated with the email you provided. The event is for members only. Please complete the application to join our community first.

Thank you for your interest in joining us

Thanks for your interest in joining the event. However, your account is still pending. Once it's approved, you'll be able to register for the event. Please contact us via help@mlops.community if you need support.

Pending Approval

Login to the Agentic AI Foundation community to connect with others, attend community events, and more!

Login or Sign Up for even more

To unlock all parts of the community and get the best experience, complete your profile.

Hey, complete your profile

It looks like there is no account associated with the email you provided. Please complete the application to join our community.

Thank you for your interest in joining Agentic AI Foundation.

Thanks for your interest in joining Agentic AI Foundation. We will review your request shortly and email you once it's approved. Meanwhile, please feel free to contact us via help@mlops.community for any questions.