The Roo Cast, the official podcast of Roo Code – Détails, épisodes et analyse

In this episode of Roo Cast, host Hannes Rudolph is joined by Roo Code founders Matt and Danny, and special guest Logan from Google DeepMind. Logan gives a deep-dive demo of the brand new "vibe coding" experience in Google's AI Studio. He showcases how to build AI-powered applications from a single prompt, using features like the "I'm feeling lucky" button to generate a meeting agenda app and even demoing image-aware models like Nano Banana live on stream.The conversation then explores the future of Google's developer ecosystem, with Logan explaining how Gemini acts as the "through line" connecting all of Google's services. The team debates how underdog tools like Roo Code can win in this new landscape and discusses the challenge of models "eating the scaffolding" that developers build. They also touch on the power of Gemini Flash Latest, the future of AI pricing, the value of code-specific models, and the path to AGI.

Resources Mentioned: Google DeepMind: https://deepmind.google/ Google AI Studio: https://aistudio.google.com/ Google Gemini: https://deepmind.google/technologies/gemini/ Google Veo: https://deepmind.google/technologies/veo/ GitHub: https://github.com/ TensorFlow: https://www.tensorflow.org/ Android Studio: https://developer.android.com/studio Google Gemma: https://ai.google.dev/gemma

In this episode of Office Hours, host Hannes Rudolph is joined by Roo Code's CTO Matt and CEO Danny, along with special guest Paige Bailey from the Google DeepMind team. Paige kicks things off with an exclusive look at some of Google's latest and most powerful AI developer products, straight from the source. She provides live, in-depth demonstrations of new capabilities, including the impressive text-to-video model Veo 3.1, and showcases the new Computer Use API by tasking a Gemini agent to navigate the web, find a specific product, and add it to a shopping cart in real-time.The conversation then shifts to the future of software development in an AI-driven world, exploring how generative tools are fundamentally changing workflows for engineers at Google, where over 50% of all production code is now AI-generated. Paige shares insights on the internal debate between model speed versus quality, the shift away from traditional product requirement documents (PRDs) toward rapid AI-assisted prototyping, and how new multimodal interfaces like the Gemini Live API are poised to redefine how developers interact with their tools. The episode wraps with a look at Google's own AI-powered code review process and a discussion on the legendary Google monorepo.

Resources Mentioned:Google DeepMind: https://deepmind.google/Google AI Studio: https://aistudio.google.com/Google AI Studio Build: https://ai.google.dev/buildGoogle Gemma: https://ai.google.dev/gemmaGoogle's Veo Model & Blog Post: https://blog.google/technology/ai/veo-updates-flow/Gemini Computer Use API Docs: https://ai.google.dev/gemini-api/docs/computer-useGemini Browser Extension: https://gemini.browserbase.com/GitHub PR with Gemini Code Assist Example: https://github.com/GoogleCloudPlatform/generative-ai/pull/2273

In this episode of Office Hours, host Hannes Rudolph is joined by Matt and Danny from the Roo Code team, along with special guests Elliot and Coston from the AI engineering consultancy, ThisDot Labs. The group dives deep into the "last mile problem" that many enterprise teams face when trying to fully integrate AI agents into their existing projects and processes. Elliot and Coston share their firsthand experiences helping clients bridge the gap between powerful AI capabilities and practical, team-wide adoption, discussing the initial skepticism they often encounter from even the most productive engineers.The conversation then shifts to a detailed breakdown of how power users and pragmatists set up their development environments, framed as a friendly "Apple user vs. Linux user" debate on customizing Roo Code. The panel also tackles a community question on the role and security of MCP servers, weighing their immense potential against the risks of third-party integrations. Finally, the team explores how AI is revolutionizing project modernization through automated documentation, highlighting the power of features like the new /init command, and discusses how AI is reshaping team structures for a future of smaller, faster, and more agile development.

In this episode of Office Hours, host Hannes Rudolph is joined by Matt Rubin, Dan, and CEO Danny, along with special guest John Sterns, CTO of Roo.vet. The team does a deep dive into the revolutionary new Cloud Agents, showcasing how the Roo.vet team has integrated them into their daily workflow. They reveal how their entire organization, including non-engineers, now uses a simple Slack integration to ask questions, fix bugs, and even prototype new features, culminating in a live demo of a product manager shipping a complete feature from a single Slack message.The conversation then shifts to managing the "PR explosion" created by this new, democratized workflow, introducing the powerful new AI-powered PR Reviewer designed to maintain code quality and streamline approvals. Hannes provides a step-by-step walkthrough on how to set up a PR Reviewer agent from scratch in the Roo Code Cloud. Finally, the team announces the official launch of the Roo Code Teams Plan, detailing the new features designed for organizational collaboration and management.

In this episode of Roo Cast, host Hannes Rudolph is joined by RooCode's lead developer Matt and CEO Danny, along with special guest Brian Fiocca from OpenAI. Brian kicks things off by sharing his extensive background in the startup world (including YC and Rescue Time) and explains his current role on the Applied Startups team at OpenAI. He then reveals the surprising story of how he first discovered Roo Code: he was looking for open-source evaluation suites to test how models like GPT-Five perform inside different coding tools and found RooCode's eval harness.The conversation dives deep into a detailed comparison of GPT-5 and GPT-5 Codex, with Brian explaining the key architectural differences and why one is more adaptable while the other is hyper-optimized for a specific harness. The team then explores the new frontier of "Evals as a Service," discussing how to move beyond simple correctness benchmarks (like Sweebench) and create "performance review" style evals for agentic tasks. The discussion also covers the critical importance of the Responses API for preserving chain-of-thought, the future of context memory, and Brian's top recommendations for the RooCode codebase.Resources Mentioned: OpenAI: https://openai.com/ Y Combinator: https://www.ycombinator.com/ Zed (Editor): https://zed.dev/ Minimax: https://minimax.chat/ Anthropic: https://www.anthropic.com/ Vercel AI SDK: https://sdk.vercel.ai/ Slack: https://slack.com/ Linear: https://linear.app/

In this episode of Roo Cast, host Hannes Rudolph is joined by Matt Rubens, Danny, and special guest JB Brown from Smartsheet. JB, one of Roo Code's top open-source contributors, shares the inside story of how Roo Code spread like "wildfire" through Smartsheet's engineering organization. He details his motivation for becoming a significant contributor, including his "selfish" reason for building the foundation of the Roo Code Marketplace: to support a private, internal-only marketplace for Smartsheet's proprietary modes.The conversation shifts to enterprise-level AI strategy, covering Smartsheet's "focus on return, not cost" philosophy for AI spend and their unique KPIs for tracking adoption. JB shares a deep dive on moving from "perishable" single-engineer prompts to a team-based AI SDLC with shared standards. This leads to a fascinating discussion on "org designing" your AI, treating modes as a specialized "workforce" of agents, and how Roo Code's new cloud agent roadmap aligns with this future of asynchronous, "digital teammates."