Sabitlenmiş Tweet
I couldn't be more excited to announce that we've raised $15M in funding, led by @FPVventures with participation from @GVteam and @QuietCapital! 🚀
English
Sagar Batchu
1.3K posts
Sagar Batchu
@sagar_batchu
CEO of @speakeasydev. Crafted tools for humans. Context for your Agents.
San Francisco, CA Katılım Mart 2022
183 Takip Edilen524 Takipçiler
We built a platform to help teams build MCP servers faster.
Then we learned building was already the easy part.
But there were problems nobody was solving:
→ No way to easily and safely provision access across your organization
→ No way to build a shared catalogue
→ Limited visibility into server usage, and identifying patterns
Wrote about what we learned and what Speakeasy MCP looks like now:
speakeasy.com/blog/we-were-w…
English
This is what we’re building at @speakeasydev. Our new MCP platform enables teams to build, secure and distribute important internal and third party tools that allows employees to super charge their ai agents - MCPs, Skills, CLIs and plugins through a single oauth experience.
Brian is 🎯
Brian Scanlan@brian_scanlan
We've been building an internal Claude Code plugin system at Intercom with 13 plugins, 100+ skills, and hooks that turn Claude into a full-stack engineering platform. Lots done, more to do. Here's a thread of some highlights.
English
Sagar Batchu retweetledi
We've been building an internal Claude Code plugin system at Intercom with 13 plugins, 100+ skills, and hooks that turn Claude into a full-stack engineering platform. Lots done, more to do. Here's a thread of some highlights.
English
The companies that invest in both end up in a compounding position.
Every well-designed internal API becomes a tool that every employee can use through Claude. The better your API layer, the more powerful your AI layer becomes.
We're a week in and just getting started.
English
Engineering teams need reliable, typed interfaces to every internal service for their coding agents. That's what the API platform builds.
Every other team needs access to those same services through AI, without understanding what's underneath. That's what the MCP platform provides.
English
A week ago we launched Speakeasy MCP Platform. Here's what I've been thinking about since. 🧵
English
When that exists, AI stops being a productivity tool for power users and starts being infrastructure for the whole company.
That's the vision behind Speakeasy MCP Platform's server catalog. One central repo. Not a local config file on someone's laptop.
English
The companies that get this right aren't just building MCP servers. They're building a catalog.
Every internal API, every third-party tool, every service the business runs on. One place. Discoverable by everyone with the right permissions.
English
Most companies experimenting with MCP today have the same problem, even if they haven't named it yet. 🧵
English
The companies that close that gap first are going to be very hard to compete with.
English
Most companies aren't bottlenecked by the AI.
They're bottlenecked by everything that sits between the AI and the tools their business runs on.
English
Fermat co-founder Shreyas Kumar was in a leadership offsite debating whether to invest eng resources in a GTM project.
While the team was still discussing, he spun up Claude agents and finished it in an hour.
The meeting didn't end. It changed subject. 🧵
English
Finally a reflection of what's happening in real enterprise. I love using CLI's with claude. It's amazing for my local dev loop and shipping/debugging production code but for big organisations with a variety of user personas MCP is often a better fit to distribute, govern and secure tool use.
English
MCPs are the opposite of dead. They are the life blood of how AI agents use services inside mid-sized and above companies.
Case in point: Uber runs on MCPs internally, for good reason. Details: newsletter.pragmaticengineer.com/p/how-uber-use…
@levelsio@levelsio
Thank god MCP is dead Just as useless of an idea as LLMs.txt was It's all dumb abstractions that AI doesn't need because AI's are as smart as humans so they can just use what was already there which is APIs
English
That's what I'm most excited about with Speakeasy MCP Platform. Not just what it enables for engineers, but what it unlocks for everyone else the moment it's deployed.
English
The bottleneck to AI adoption at most companies isn't the AI. It's everything that has to happen before someone can actually use it.
Unified auth takes that off the table.
English
Think about what day one looks like at most companies.
A new hire spends the first week getting access to things. Tickets to IT. Waiting on approvals. Twelve different UIs. By the time they're productive, two weeks are gone. 🧵
English