HomeCompaniesRindler

The IDE for Tokenmaxxers

We build Maxxwell, a desktop app for people who run several coding agents at once. Past two or three sessions, you become the bottleneck. Every session asks you something, most of it small, and you spend the day answering instead of deciding. It is also hard to tell which sessions are making progress, which are stuck on something trivial, and which have quietly started building the wrong thing. Maxxwell puts every session in one window and marks each one working, idle, or waiting on you. On top of that sits an orchestrator, a real coding-agent session of its own that reads the others and reports back. You talk to it instead of to twelve terminals. It tells you what landed, what it decided for you, and what actually needs your call. Every worker is your own tool, unmodified. Claude Code, Codex and Cursor Agent run in real terminal sessions you can attach to and take over at any point
Active Founders
Michael Serrano
Michael Serrano
Founder/CEO
CEO of Rindler. MIT SB + MEng in Physics and Computer Science. Prev: ML engineer at Roblox, LLM research at MIT CSAIL.
Arthur De Los Santos
Arthur De Los Santos
Founder/CTO
MIT '26 in CS/AI/ML We're building Rindler, the infrastructure layer that allows AI agents to buy, book, and browse on any website.
Company Launches
Rindler - Turn Any Website Into An API For AI Agents
See original launch post

TL;DR: Rindler turns websites into deterministic APIs for AI agents. We map a site once into screens, actions, and structured outputs, so agents can read data and take actions quickly and reliably.

Ask: If you’re building an agent that needs to browse the web, compare products, fill forms, search listings, use authenticated portals, or act on third-party websites without an API, we’d love to help.

https://youtu.be/V_nCIxYD-so

Problem

More products are starting to depend on AI agents that can do real work on the web: shop across retailers, pull housing listings, use job boards, submit forms, monitor prices, update back-office systems, and operate inside customer portals.

But most browser agents still treat every website visit like a new puzzle. On every run, they need to decipher the page, infer what matters, decide what to click, and recover when the page behaves differently.

That works for demos, but it breaks down in production. Real websites have pop-ups, login gates, cookie banners, bot defenses, slow-loading forms, dynamic filters, fragile dropdowns, layout changes, and pages that look similar but behave differently.

The result is slow execution, high token cost, and workflows that are hard to trust.

Solution

Rindler maps websites into deterministic APIs for agents.

Instead of forcing an agent to re-explore a site every time, Rindler captures the site as structured screens, available actions, and typed outputs. Agents can then call actions like search, add_to_cart, or download_records, depending on the mapped site.

At runtime, Rindler handles the browser work through a remote MCP endpoint:

  • Returning structured data
  • Applying dynamic filters
  • Managing popups, bot defenses, and layout changes
  • Supporting authenticated flows

Background

We started Rindler because “AI agent uses a browser” is compelling, but the current browser-agent loop is too brittle for teams building real products.

If an agent needs to use the same third-party site hundreds or thousands of times, it should not have to rediscover that site from scratch on every run.

Rindler is still a work in progress. Some sites map cleanly, others require more in-depth mappings, and bot-defended or logged-in sites may need a one-time credential or cookie capture.

We’re looking for feedback from founders building products where reliable web actions are part of the core workflow.

If there’s a website you wish your agent could use reliably, send it to us!

uploaded image

YC Summer 2026 Application Video
Rindler
Founded:2025
Batch:Summer 2026
Team Size:2
Status:
Active
Location:San Francisco
Primary Partner:Ankit Gupta