Automate Web3 Trading Strategies

Design and deploy fully-automated,
permissionless trading strategies on Web3 in minutes. No coding required.

Execute on proven
venues

Strategies receive trading-only permissions on the venue you choose. Funds never leave your vault. Go live on Drift today, with Hyperliquid, Jupiter, and pump.fun next.

Drift
Hyperliquid
Jupiter
Pump.fun

Readable Strategies, Predictable Outcomes

Strategies conform to an opinionated structure that maximizes clarity and guarantees deterministic execution.

I'd like a delta-neutral strategy aimed at Bitcoin

avatar
logo

Here's the classic funding rate arb parameterized with a minimum net funding rate delta to make sure you aren't overtrading and losing due to trading fees.

Strategy visualization

Autonomous Trading.
Purely Permissionless.

Instantly adapt and execute automated strategies while your assets remain safely
non-custodial.

  • A portfolio that reacts to the market
    A portfolio that reacts to the market

    Signal values determine strategy state and state transitions with simple if/else control flow.

  • Permissionless by Design
    Permissionless by Design

    Wallet Authentication. Strategies get trading only permissions; No accounts. No emails. No KYC.

  • More Automation
    More Automation

    Structure executes your trading strategy, making trades and rebalancing automatically.

Tick-Level Data Driving
Real-Time Trading
Decisions

Signal values are computed from tick-level market data yielding strategy state transitions that are indicated by simple if/else control flow.

Tick Data

An anatomy that maximizes clarity
enabling easy reasoning.

  • Data, Computations, & Signals

    Turn tick-level market data into deterministic signals; every value is reproducible and auditable.

    Data, Computations, & Signals

  • State Machine

    Signal values determine strategy state and state transitions with simple if/else control flow.

    State Machine

  • Target Positions

    Each state resolves to one set of target positions. An executor manages orders to achieve those target positions outside of the strategy logic.

    Target Positions

Learn & Adapt

Backtest on Structure, side-by-side with benchmarks or other strategies. Fees, slippage, and final value are calculated for you.

Learn & Adapt

Frequently Asked Questions

How does Structure’s AI turn a natural‑language idea into a live trading strategy? +
Describe your idea and our AI generates a Strategy DSL that defines signals, conditions, states, and target positions. You review or edit the draft, then we compile it to Rust and validate it using the same engine that runs in production. Run a tick‑level backtest and, once satisfied, deploy to production.
Can I use MCP agents or code to build and test instead of the UI? +
Yes. You can author the DSL in code, keep it in Git, and use our API/CLI to compile, backtest, and deploy. MCP agents can call the same actions to generate or modify the DSL and run tests, so the UI is optional.
Which DEX venues and chains are supported at launch? +
At launch, Structure supports all Drift markets on Solana. Additional venues and chains are on the roadmap.
How are signals created from tick‑level data, and how do they drive state changes? +
Strategies on Structure receive real‑time tick updates for all required market data. On each event, the strategy recomputes its derived features and signals to determine the next action. If a state change is triggered, the strategy’s target positions are updated and executed by the platform’s target‑position executor. This design ensures fast, reliable reactions to changing markets.
What’s the difference between data, computations, signals, states, and target positions? +
Data are external market events such as order‑book updates, trades, and funding rate changes. Computations are mathematical transforms of data (and prior computations) that produce features. Signals are boolean indicators derived from these features that determine whether the strategy should transition into a new state. A state is a named mode that prescribes a target position—the desired long or short exposure per instrument—which the executor achieves via orders.
How do backtests work and what metrics can I expect to see? +
Backtests replay tick‑level data over the selected period through the same code that runs in production, including fees, funding, slippage, latency, and partial fills. You can run single tests, walk‑forward splits, and parameter sweeps. Metrics include PnL, max drawdown, Sharpe/Sortino, fee and funding breakdowns, execution slippage, state timelines, and per‑period or per‑market performance.
How are credits billed in USDC, and what actions consume them? +
Users pre‑purchase credits in USDC; credits are consumed based on the compute consumed by testing and production runs. On‑chain gas fees are included in this cost. Each plan includes a defined resource allotment, and you can top up credits at any time for additional usage.
What safeguards prevent unauthorized withdrawals or trades? +
You grant the strategy trade‑only permissions on the connected DEX account; custody remains with the DEX program and your wallet. Withdrawal permissions are not granted to Structure strategies. Granting withdrawal rights out of band is strongly discouraged.
How do I version, review, and roll back strategy changes? +
Every edit creates a new, immutable strategy revision that includes the strategy graph, parameters, and compiled build. In the UI, you can navigate backward and forward across revisions, review changes, and choose which one to run. Rollbacks are explicit and fully logged.
Do you currently take any trading fees on top of what the exchange charges me? +
No. Structure does not charge trading fees on top of the exchange’s fees.

Backed by the
investors behind
Coinbase and Solana,
founded by
MIT and Jump Trading
alumni.

Get Early Access
To Our Launch