
22. building cloud infra @maritime_sh (yc f26) creating community @disruptors_bu
San Francisco, CA
Joined June 2024
- Tweets529
- Following720
- Followers802
- Likes1.1K
Ary Shrivastava retweeted
We gave AI its own desktop on your computer
Today we're launching River (YC F26)! River can use your apps and browser, read your files, and even control your phone. It does all of that without taking over your screen
For example, I can ask River to plan a two-day onsite, then get back to coding while it checks my calendar, books flights, and sends the invites
River’s handled over 1.25M requests from 100,000+ users. Setup takes two minutes. Give it a try for free at meetriver.com
Ary Shrivastava retweeted
⚓ @maritime_sh (YC F26) gives every AI agent its own isolated persistent computer. More and more startups are building agent products where each customer gets their own dedicated agent. That means running thousands of agents where they're all on a machine that keeps its files and logins in between tasks. @mariagorskikh and @AryShriv experienced this firsthand while hosting hundreds of agents during their research, where tons of time went into keeping envs running and debugging individual agents.
Maritime takes all of that off your plate. Agents don't even have to move, they stay on their existing backend and control a Maritime computer over MCP. As we head towards a world of agents, someone has to run the hardware that they live on. Super pumped to support Maria and Ary, congrats from @pioneer_fund!
Ary Shrivastava retweeted
Everyone is building sandboxes.
But a sandbox is only 10% of the infrastructure you need to run AI agents. You're left building the other 90% in-house: deployment, persistent state, monitoring, recovery, scaling
You're trying to sell agents. Somehow you're building a cloud.
Today we're launching Maritime out of YC F26 🚢
Maritime is a managed cloud for AI agents. We handle the infrastructure so you can focus on your product.
a guy at a party told me his team tried to reverse engineer maritime and failed
if people are trying to reverse engineer your product you know you’re doing something right
Ary Shrivastava retweeted
Epsilab is working with frontier labs to build RL environments for long-horizon enterprise work.
Today, low-quality data is an industry-wide problem. We founded Epsilab to bring together real-life data and top 1% domain expertise to solve this.
Backed by @ycombinator (F26).
Epsilab Launch
YC F26. Frontier intelligent for RL
Introducing Apple Messages for Business on @onkeiki 🍏
You can now connect your agent to Apple Messages for Business, so customers can text you from Apple Maps, Safari or your site and get answers under your verified name and logo.
Early access starts today 👇
Massive congrats to Letter for their public launch! Excited to be supporting their new Letterpress-1 beta with @maritime_sh
Impeccable product and team :)
Today, the @lettercompany is releasing the beta of Letterpress-1, the world’s first coding agent, harness, and IDE purpose-built for growth engineering.
tl;dr: check out the 🧵 to see the product in action.
Growth is re-emerging as an engineering problem, but today’s coding tools weren’t designed for it. Your website isn’t isolated like your app: it reacts to the outside world (e.g. agent behavior, Google rankings, observed human interactions, and competitor activity).
Leading coding agents were built to scale complex codebases, not to autonomously build for outside signals. Letterpress-1 is.
Letterpress-1 is designed to tackle growth problems of every size. It’s also proactive: it suggests the nth next thing to build for growth. It'll issue pull requests for de-anonymization, experimental landing pages, warmed emails, blog automations, agent telemetry, etc.
The Letterpress IDE is also built into your site. Press ⌘+L, authenticate, and edit text directly on your website. Letterpress-1 automatically opens a pull request with the change.
Letterpress-1 ships with readily available infrastructure to stage futuristic growth strategies that win over agents. These include:
=> Proofs: autonomous evidence chains that verify customer success data for agent visitors
=> Simulations: staged mirror applications where agents can safely try solutions before buying
=> Benches: self-hosted, reproducible benchmarks with open-source, independent governance
=> Phantoms: autonomous websites to sandbox growth strategies
Build, automate, and refine your growth process with the Letterpress family of tools. We're proud to count some of the world's most forward-thinking startups as our customers.
Sign up today at letter.dev.
I’ll share a few examples of Letterpress-1 in action in the thread below.
9
530
Ary Shrivastava retweeted
Latest Maritime changelog:
- Ephemeral sandboxes from $0.022 an hour
- Swarms: one job across 1,000 sandboxes
- Windows desktops through MCP
we’re teaming up with @supermemory @onkeiki @pumpgtm to help you go viral!
the two things that’ll kill you as a b2c agent startup is bad infra and terrible distribution, and this is the stack that will let you build a truly useful, monetizable agent like instinct
go join the races at competition.maritime.sh
Ary Shrivastava retweeted
I don’t want muse to rule the world.
The next generation of agent companies is being started right now, by people shipping agents that real users text every day.
If that’s you, build it now. Keiki and Maritime cover the tools, hosting and LLM credits. Get to 1,000 users and win $20k in credits plus a trip to SF.
competition.maritime.sh/
we’re teaming up with @onkeiki to give you the opportunity to build the NEXT INSTINCT
we’re living in a gold rush, especially when it comes to personal agents. but the moat is no longer technical ability, it’s distribution and execution
if you get 1000 real people using your agent, we’ll give you $20k in maritime llm credits and will fly you out to sf
go apply at competition.maritime.sh
Ary Shrivastava retweeted
AI personal assistants are the next gold rush.
And we want OUR users to build the next unicorns.
So @maritime_sh is partnering with @onkeiki to give you everything you need to build and scale the next Instinct.
The first teams to reach 1,000 users get $20,000 in LLM credits and a 3-day trip to SF.
Go build: competition.maritime.sh
Ary Shrivastava retweeted
I sat down last week with @mariagorskikh from @maritime_sh (YC F26) to discuss the future of sandboxes and agent infrastructure.
This is my second time doing YC, and one of my goals this batch was to not just highlight ourselves, but some of the other amazing companies in the batch.
This conversation is the first of many to come. Looking forward to what @mariagorskikh and @AryShriv build.
Day 6 of 6 of #SFTechWeek, and it sure flew by.
We started the morning building AI teams at our hands-on workshop with Turnstone (W26) and UpScaleX, then wrapped the week at @maritime_sh (YC F26) launch party at Corgi Cafe with agents, corgis, cocktails and a line out the door.
We enjoyed every moment of the chaos. See you next year, same time, same place 🧡
















