Artwork

内容由Damien Filiatrault and Scalable Path提供。所有播客内容(包括剧集、图形和播客描述)均由 Damien Filiatrault and Scalable Path 或其播客平台合作伙伴直接上传和提供。如果您认为有人在未经您许可的情况下使用您的受版权保护的作品,您可以按照此处概述的流程进行操作https://zh.player.fm/legal
Player FM -播客应用
使用Player FM应用程序离线!

Unicorn Lessons from RapidAPI’s Founder: A Conversation with Iddo Gino

37:59
 
分享
 

Manage episode 512891465 series 3669449
内容由Damien Filiatrault and Scalable Path提供。所有播客内容(包括剧集、图形和播客描述)均由 Damien Filiatrault and Scalable Path 或其播客平台合作伙伴直接上传和提供。如果您认为有人在未经您许可的情况下使用您的受版权保护的作品,您可以按照此处概述的流程进行操作https://zh.player.fm/legal
Host Damien Filiatrault welcomes Iddo Gino, founder of RapidAPI (launched at 17, later a unicorn, acquired by Nokia) and now CEO of Datawizz. They trace Rapid’s journey from a GitHub list to a global API marketplace, unpack what sky-high valuations actually change, and debate MCP versus simply fixing the APIs we already have. Iddo then shares how Datawizz slashes LLM bills with tiny, task-specific models and why the future leans hard toward edge and on-device inference.
What you’ll learn
  • How “Awesome APIs” (a GitHub repo) became RapidAPI’s interactive playground, marketplace, and eight-year scaling story, ending in a Nokia acquisition
  • What high valuations really buy (credibility, recruiting) and why raising “too much” is a double-edged sword for efficiency
  • A contrarian take on Anthropic’s Model Context Protocol (MCP): why reinventing interfaces may just create a second integration surface to maintain
  • The boring fix that beats new protocols: clean REST/GraphQL, accurate OpenAPI specs, and docs that match reality
  • Datawizz’s playbook: route requests to small, specialized models (and only fall back to big LLMs when needed) for 85–95% cost reductions
  • How the router works in practice (OpenAI-compatible endpoint), when volume justifies specialization, and why clustering real traffic matters
  • Edge & on-device AI: running custom adapters on iOS’s built-in models, Chrome’s Gemini Nano, and at the edge on Cloudflare AI Workers for latency, privacy, and cost wins

Memorable sound bites
  • “V1 of Rapid was just a GitHub repo called Awesome APIs.”
  • “If LLMs can’t use your API, your developers probably can’t either, fix the API.”
  • “MCP feels like reinventing APIs. In a few years it could be just as messy, and now you’re maintaining two surfaces.”
  • “Our customers see 85–95% cost reduction by routing to small, specialized models.”
  • “On-device is free, fast, and private.”

Tune in for a founder’s-eye view of scaling an API marketplace, a pragmatic critique of shiny new protocols, and a concrete roadmap to cheaper, faster AI through specialization and the edge.

Get 20% off your first month with Scalable Path: https://www.scalablepath.com/commit
Commit & Push Website: https://www.commit-push.com/
Scalable Path Website: https://www.scalablepath.com/
  continue reading

8集单集

Artwork
icon分享
 
Manage episode 512891465 series 3669449
内容由Damien Filiatrault and Scalable Path提供。所有播客内容(包括剧集、图形和播客描述)均由 Damien Filiatrault and Scalable Path 或其播客平台合作伙伴直接上传和提供。如果您认为有人在未经您许可的情况下使用您的受版权保护的作品,您可以按照此处概述的流程进行操作https://zh.player.fm/legal
Host Damien Filiatrault welcomes Iddo Gino, founder of RapidAPI (launched at 17, later a unicorn, acquired by Nokia) and now CEO of Datawizz. They trace Rapid’s journey from a GitHub list to a global API marketplace, unpack what sky-high valuations actually change, and debate MCP versus simply fixing the APIs we already have. Iddo then shares how Datawizz slashes LLM bills with tiny, task-specific models and why the future leans hard toward edge and on-device inference.
What you’ll learn
  • How “Awesome APIs” (a GitHub repo) became RapidAPI’s interactive playground, marketplace, and eight-year scaling story, ending in a Nokia acquisition
  • What high valuations really buy (credibility, recruiting) and why raising “too much” is a double-edged sword for efficiency
  • A contrarian take on Anthropic’s Model Context Protocol (MCP): why reinventing interfaces may just create a second integration surface to maintain
  • The boring fix that beats new protocols: clean REST/GraphQL, accurate OpenAPI specs, and docs that match reality
  • Datawizz’s playbook: route requests to small, specialized models (and only fall back to big LLMs when needed) for 85–95% cost reductions
  • How the router works in practice (OpenAI-compatible endpoint), when volume justifies specialization, and why clustering real traffic matters
  • Edge & on-device AI: running custom adapters on iOS’s built-in models, Chrome’s Gemini Nano, and at the edge on Cloudflare AI Workers for latency, privacy, and cost wins

Memorable sound bites
  • “V1 of Rapid was just a GitHub repo called Awesome APIs.”
  • “If LLMs can’t use your API, your developers probably can’t either, fix the API.”
  • “MCP feels like reinventing APIs. In a few years it could be just as messy, and now you’re maintaining two surfaces.”
  • “Our customers see 85–95% cost reduction by routing to small, specialized models.”
  • “On-device is free, fast, and private.”

Tune in for a founder’s-eye view of scaling an API marketplace, a pragmatic critique of shiny new protocols, and a concrete roadmap to cheaper, faster AI through specialization and the edge.

Get 20% off your first month with Scalable Path: https://www.scalablepath.com/commit
Commit & Push Website: https://www.commit-push.com/
Scalable Path Website: https://www.scalablepath.com/
  continue reading

8集单集

すべてのエピソード

×
 
Loading …

欢迎使用Player FM

Player FM正在网上搜索高质量的播客,以便您现在享受。它是最好的播客应用程序,适用于安卓、iPhone和网络。注册以跨设备同步订阅。

 

快速参考指南

版权2025 | 隐私政策 | 服务条款 | | 版权
边探索边听这个节目
播放