Skip to content
Bavariance

What we have built

PolySimulator, the market data behind it, and the agent system that builds and runs both.

PolySimulator

A simulation platform for event markets. Traders practice on live Polymarket markets with paper money; trading software runs against the same markets through an API that follows Polymarket's official client. Orders fill level by level against the real order book, with the venue's fees.

About PolySimulatorPython SDK on GitHub

Live since
November 2025; first account 30 November 2025
Registered accounts
7,000+, of which 3,700+ have placed a trade
Simulated trades
390,000+ since launch
API
80 accounts with an API key; Python SDK on PyPI since June 2026

PolySimulator production database, 16 October 2026. SDK release dates from PyPI.

PolySimulator's market page: the live order book and the trade ticket, with the fee and slippage a paper order would pay
PolySimulator, 16 October 2026: a market's live order book and the trade ticket, with the fee and slippage a paper order would pay.

Engineering and operations run by agents

PolySimulator's code is written by AI coding agents, and most changes get an AI review before they merge. Agents merge into our working branch and run staging themselves; in production they act under a standing permission from us. They also run the day-to-day operations.

How the agents are organized
Merged pull requests
2,699 in 90 days, 980 of them in the last 30GitHub, PolySimulator's main repository, 16 October 2026.
Availability
API 99.98% since 9 August 2026, measured by two independent probes. Not an SLA; see the status page.PolySimulator status page, 16 October 2026.
Lessons
188 written lessons in a shared memory of 1,686 notes, synced across machines every 20 minutesOur agent memory, 16 October 2026.

  • Watching

    Agents watch our servers, the data capture and the public status of every service, and open a ticket when an alert fires.

  • Housekeeping

    Backups, data reconciliation and clean-up jobs run on schedules the agents set up and check.

  • Briefings

    Every morning and evening we get a briefing on what changed and what needs a decision from a person.

How changes are checked

Agents write the code, so the checks matter. These run today; our rules also require a signed-in browser test after each app change and a failing test before each fix.

  • AI review before merging

    In the 30 days to 16 October 2026, 459 of the 768 pull requests that changed application code got an AI review verdict before merging that names the commit it checked; for 340 of them, that was the exact commit that merged. 208 were sent back at least once. The reviewers ran on models from Anthropic, OpenAI, Google and others.

  • Contract checks

    Pull requests run automated checks for API-schema drift, route changes, pricing consistency, leaked secrets and oversized diffs.

  • A guarded production branch

    Merging into the production branch needs a person's approval after the last push, and three required checks.

  • Bugs become rules

    When a leaderboard cache failed because of a database timeout set for a whole session, a lint rule began flagging that pattern in new code. Our rules require the same for every repeated defect.

  • A trade every 20 minutes

    A synthetic trade runs against production about every 20 minutes, 497 times in the last week, to prove that orders still go through.

Pull requests merged into our working branch, 16 September to 16 October 2026; production probe log.

Our market data

Since 30 July 2026 we record Polymarket's full order-book feed ourselves: on the order of 100 million updates an hour, tens of terabytes so far, with settlement reference prices alongside. It is what our research on fill fidelity and point-in-time replay runs on.

The update rate is approximate, measured in August 2026; the archive size is from October 2026.

Our research

Public code

  • polysimulator-sdk

    The official Python SDK for PolySimulator's paper-trading API, with a drop-in client for code written against Polymarket.

  • polysimulator-docs

    The API documentation, synced from the main repository and readable by AI coding tools.

  • polysim-bootstrap

    The one-command setup for our development machines: toolchain, AI coding agents and their tool servers, with secrets pulled from a vault and none kept in the repository.

  • super-board, extended

    A build, test and review loop for coding agents, driven by a project board. It is Eric Tech's open-source tool; Wladimir extended it with builder and tester agents on models from other providers, a binding review by a second model family, and ideas broken down into detailed tasks before any agent starts.

Want an agent system like this for your company?

Write to contact@bavariance.com. A few lines on the process you have in mind are enough to start. We reply in German or English.