Design partner program now open for funds and quant desksDesign partner program now open
Ensemble

The index layer for prediction markets.

Ensemble gives hedge funds, quant desks and active traders rules-based baskets of Kalshi and Polymarket US contracts, with fee-aware backtests and execution that runs inside your own infrastructure.

  • Self-hosted or desktop
  • No custody of funds
  • Every trade approved
Index monitorSimulated
Fed PathTheme100.00
ContractVenueYesWeightChg
Simulated data for illustration, not live prices
$120B

Notional prediction market volume across nine venues, 30 days to Oct 29, 2026

$240B

Projected sector volume for 2026, from Bernstein

$40B

Kalshi valuation after its October 2026 funding round

1.2M

Active traders on Kalshi at the end of 2025

Sources: DeFi Rate, The Block (Bernstein), KFGO, Sacra. Notional volume counts $1 per contract.

Why an index layer

Thousands of markets. No portfolio tools.

Event contracts now trade at the scale of a real asset class, but every venue and terminal still treats them one ticket at a time. Ensemble treats them as a book: grouped, weighted, rebalanced and measured.

We don't sell alpha. A basket of fairly priced contracts still pays its fees. What a desk gets is the machinery around the trade: hedges that line up with real exposures, systematic sleeves that run on rules, and execution that keeps fees and slippage in check.

Thirty contracts and the index that holds them Each vertical bar is one contract split into its chance of yes and no. A white line runs through them as their smoothed average, steadier than any single bar.
Each bar is one contract: chance of yes, chance of no. The line is the index: the average of all of them, steadier than any one.
Why desks switch

The same trades, without the spreadsheet

Most desks trading event contracts today run baskets by hand: a spreadsheet of tickers, one order ticket per leg and a screenshot for the audit file.

TaskBy handWith Ensemble
Building a basketPick contracts from each venue's site and size them in a spreadsheetA versioned rulebook screens every market and sizes each leg
Knowing the costFees and slippage show up after the fillsEvery leg is quoted with its exchange fee before you approve
Entering the tradeOne ticket per leg, across two venues, while prices moveOne approval sends every leg as limit orders with a slippage cap
RebalancingNoticed late, if at allDrift and eligibility triggers propose trades, skipping small ones
Settlements and disputesTracked in a calendar and chased manuallyHandled by written rules: rolled, frozen or flagged
Audit and reconciliationScreenshots and exported statementsEvery quote, approval and fill logged; ledger checked every 5 minutes
Solutions

Built for the desks that trade event risk

Three ways teams use Ensemble, from hedging a book to trading a single release.

Hedge funds and asset managers

Self-hosted edition

Offset rate, inflation, crypto and weather exposure in an existing book with baskets that pay when the event hits.

  • Hedge ratios sized to a target notional
  • Two-person approval and audit log
  • Runs in your cloud, with your SSO
Self-hosted edition

Quant and prop desks

Self-hosted edition

Run systematic sleeves across hundreds of markets from rulebooks in your own repo, tested against real fees and depth.

  • Python SDK and REST API
  • Backtests on historical order books
  • Maker-first execution with slippage caps
See the platform

Active and day traders

Desktop edition

Enter and exit a whole basket around a single event, like every contract riding on a CPI print, in one order.

  • One-click basket entry and exit
  • Build and backtest your own baskets
  • Keys stay in your keychain
Desktop edition
Worked example

Hedging a rates book into a Fed meeting

A macro fund is positioned for rate cuts. If the Fed holds and inflation runs hot, the book loses. Event contracts can pay exactly in that scenario.

  1. Write the hedge as a rulebook.Three legs across two venues: no December cut, no January cut, and a hot core CPI print. Hedge ratio set to a $260,000 maximum payout.
  2. Check the cost before trading.Ensemble quotes every leg with its exchange fee. Resting as a maker keeps fees to about $756 on $120,400 of contracts.
  3. Approve once.One person proposes, a second signs off, and all three legs go out as limit orders.
  4. Let the rules run it.As legs settle or drift, Ensemble proposes the roll or rebalance for approval.
Hedge basketIllustration, simulated prices
  • Fed cuts at the December meetingKalshi · buy No at 0.38
    120,000$45,600
  • Fed cuts at the January meetingPolymarket US · buy No at 0.59
    80,000$47,200
  • Core CPI 0.3% or higher in NovemberKalshi · buy Yes at 0.46
    60,000$27,600
Contracts
$120,400.00
Exchange fees, maker
$755.58
All-in cost
$121,155.58

What the basket pays

  • Holds in December and January, CPI hot$260,000+$138,844
  • Cuts in December, holds in January, CPI hot$140,000+$18,844
  • Holds in December, cuts in January, CPI soft$120,000−$1,156
  • Cuts at both meetings, CPI soft$0−$121,156
Like any hedge, the basket costs money in the scenario the book is positioned for. Prices and sizes are illustrative, not a recommendation.
Platform

From rulebook to fills

Every index is a versioned rulebook, so the backtest, the live basket and the audit trail can never disagree.

Full methodology
rulebooks/fed-path.yamlRunning now
id: fed-path
version: 1.0.0
kind: theme
universe:
  - venue: kalshi
    series: KXFEDDECISION
    events: 3              # next three meetings
    title_match: maintains
    side: "no"
filters:
  price_band: [0.03, 0.97]
  max_spread: 0.05
  min_open_interest: 1000
  min_hours_to_close: 24
weighting: risk_parity
group_cap: 0.5             # per meeting
rebalance:
  every: weekly
  drift_trigger: 0.05
Indexes

Reference indexes

Five published rulebooks across economics, crypto, weather and elections, each recalculated every 15 minutes from live Kalshi and Polymarket US prices. No sports at launch. Fork any of them or write your own.

How indexes are built
IndexTypeUniverseWeightingRebalanceVenuesLevel
Fed Path
ThemeNo on “Fed maintains rate” for each of the next three FOMC meetingsRisk parity, 50% cap per meetingWeekly, and rolls a day before each meetingKalshiView
Inflation and Jobs
ThemeThe contract nearest the median for the next CPI, core CPI, payrolls, unemployment and jobless claims releasesRisk parity, 35% cap per releaseWeekly, and rolls with each releaseKalshiView
Crypto Downside
HedgeYes on Bitcoin and Ether falling below set prices this month and this yearEqual payout, 40% cap per seriesWeeklyKalshi, Polymarket USView
Weather Shield
HedgeYes on monthly snowfall above a set depth in New York, Boston, Chicago and DenverEqual payout, 35% cap per cityMonthly, and rolls each monthKalshiView
Election 2028
ThemeYes on the current favorite for the winning party and both nominationsRisk parity, 50% cap per raceMonthlyKalshiView

Levels are paper portfolios that started at 100 on November 2, 2026 and pay real spreads and exchange fees on every trade after launch. They are not investable products. Strategy sleeves built on documented pricing patterns, such as the favorite-longshot bias, are published only after they clear their fees and spreads in testing.

Venues and integrations

Two regulated venues, one book

Ensemble connects to your own accounts on both CFTC-regulated venues and fits into the tools your team already runs.

Kalshi

Connection
Your firm's API keys, or orders through your existing FCM
Authentication
Ed25519 or RSA-PSS signed requests
Orders
Limit orders, batch create and cancel, order groups as a kill switch
Taker fee
0.07 × p × (1 − p) per contract, by series
Maker fee
0.0175 × p × (1 − p) per contract
Testing
Full sandbox against Kalshi's demo exchange

Polymarket US

Connection
Your firm's Polymarket US API keys
Authentication
Ed25519 signed requests
Orders
Limit and market; day, good-till-cancelled, immediate-or-cancel and fill-or-kill
Taker fee
0.0695 × p × (1 − p) per contract
Maker fee
None
Testing
Paper mode on live prices before any order is sent

Your stack

  • Python SDK and REST APIDefine rulebooks, run backtests and pull positions from your own code
  • Single sign-onSAML or OIDC, with roles for proposers, approvers and viewers
  • SecretsAWS Secrets Manager, Google Secret Manager, HashiCorp Vault or the OS keychain
  • DeploymentA container image for your cloud or data center
  • ExportsPositions, fills and the audit log as CSV or JSON

Fee formulas are each exchange's published schedule as of November 1, 2026; p is the contract price in dollars. Ensemble reads the current schedule before every quote.

Security

Your keys never leave your environment

Ensemble's servers publish rulebooks and market data. Orders travel directly from your infrastructure to your exchange accounts.

No custodyEnsemble never holds funds or sees keys
Approval policiesOne- or two-person sign-off per index
Audit logEvery quote, approval and order, exportable
ReconciliationLedger checked against each venue every 5 minutes
Kill switchesGlobal, per index and per venue
Eligibility checksState availability and excluded markets, before every order
Regulatory posture

Software, not a market participant

Ensemble is built so your firm stays the only party that holds positions and decides trades.

Regulated venues onlyKalshi and Polymarket US are CFTC-regulated exchanges; Ensemble trades nowhere else for US accounts
No custodyFunds and positions stay in your exchange accounts; Ensemble never holds money
No discretionNothing executes without an approval your team controls, and every approval is recorded
Published rulesReference indexes follow public, versioned rulebooks, the same for every user
Eligibility before every orderState availability and excluded categories are checked first
Records you ownQuotes, approvals, orders and fills are exportable for your compliance team

This is a summary of how Ensemble is designed, not legal advice. Firms should confirm their own obligations before trading event contracts.

Editions and pricing

One engine, two editions

Teams license the self-hosted engine annually. Desktop is a monthly subscription.

Feature Self-hostedFunds, prop and quant desksAnnualRequest a demo DesktopActive and day traders$49 a monthJoin the waitlist ResearchPaper trading demo$0Try the demo
Indexes
Reference rulebooks and paper levelsIncludedIncluded
Custom baskets and rulebooksIncludedNot included
Backtests with real fees and depthIncludedNot included
Execution
Basket execution on Kalshi and Polymarket USIncludedNot included
Maker-first orders and slippage capsIncludedNot included
Python SDK and REST APINot includedNot included
Controls
Two-person approval, SSO and rolesNot includedNot included
Audit log exportIncludedNot included
Runs in your cloud or data centerNot includedNot included

Desktop

Active and day traders

$49 a month

Join the waitlist
  • Indexes
  • IncludedReference rulebooks and paper levels
  • IncludedCustom baskets and rulebooks
  • IncludedBacktests with real fees and depth
  • Execution
  • IncludedBasket execution on Kalshi and Polymarket US
  • IncludedMaker-first orders and slippage caps
  • Not includedPython SDK and REST API
  • Controls
  • Not includedTwo-person approval, SSO and roles
  • IncludedAudit log export
  • Not includedRuns in your cloud or data center

Research

Paper trading demo

$0

Try the demo
  • Indexes
  • IncludedReference rulebooks and paper levels
  • Not includedCustom baskets and rulebooks
  • Not includedBacktests with real fees and depth
  • Execution
  • Not includedBasket execution on Kalshi and Polymarket US
  • Not includedMaker-first orders and slippage caps
  • Not includedPython SDK and REST API
  • Controls
  • Not includedTwo-person approval, SSO and roles
  • Not includedAudit log export
  • Not includedRuns in your cloud or data center

Exchange fees are charged by Kalshi and Polymarket US and shown before every trade. Early-access pricing may change before launch; design partners receive preferred terms.

Early access

Request a demo

We're onboarding a small group of design-partner firms first, then opening Desktop to the waitlist.

  1. Intro call, 30 minutesYour use case, venues and the indexes you'd trade.
  2. Sandbox, week 1Ensemble wired to Kalshi's demo exchange and paper mode on Polymarket US, so you test without real money.
  3. Deploy, weeks 2-4The container in your cloud with your SSO and secrets manager, running in paper mode against live prices.
  4. Go live with limitsSmall size caps and two-person approval first, then your own limits. Design partners get preferred pricing and a say in the roadmap.
Edition

We'll only contact you about Ensemble. See our privacy policy.

FAQ

Questions from desks

Something else? hello@ensemble.trade

Is Ensemble a fund or a broker?

Neither. Ensemble is software licensed to you. Your firm holds every contract in its own exchange accounts, and Ensemble never holds money, routes orders through its own systems or sells shares of anything.

Can we run it inside our own cloud?

Yes. Self-hosted ships as a container for your cloud or data center, uses your secrets manager and single sign-on, and only reaches out to Ensemble for rulebooks, market data, license checks and updates.

How are backtests built?

On historical order books, with each venue's published fee schedule and fills limited to visible depth. Frictionless backtests overstate returns on thin, fee-heavy markets, so we don't use them.

How much capital can an index absorb?

Many event markets are thin. Every rulebook sets per-leg size caps tied to visible depth, and each index publishes its capacity so you know where your own flow starts moving prices.

Does it trade automatically?

Nothing executes without an approval step your team controls. Self-hosted supports two-person approval, so one person proposes and another signs off.

Is it worth it for a day trader?

If you trade one market at a time, probably not. If you trade around events like a jobs report or a Fed meeting, Desktop lets you enter and exit a basket of related contracts in one order with a slippage cap.

Which exchanges and states?

Kalshi and Polymarket US, for US accounts in the states where each operates. Ensemble checks availability before every order. No sports contracts at launch.

Is Ensemble affiliated with Kalshi or Polymarket?

No. Ensemble is independent software that connects to your own accounts through each exchange's public API.

What does it cost?

Self-hosted is an annual license, with preferred terms for the first design-partner firms. Desktop is $49 a month. The paper-trading demo and published reference indexes are free. Exchange fees are charged by each exchange and shown before every trade.

How does Ensemble connect to Kalshi if we already clear through an FCM?

Self-hosted can route Kalshi orders through your existing futures commission merchant or use your firm's own Kalshi API keys, whichever your operations team prefers. Polymarket US connects with your own API keys.

Can we write our own rulebooks?

Yes. Rulebooks are YAML files that live in your own repository, so they go through your normal code review. Each change is a new version, and backtests record which version they used.

What leaves our environment?

License checks, update downloads and requests for rulebooks and historical market data. Keys, balances, positions, orders and approvals stay on your systems.

What happens if an exchange goes down mid-rebalance?

Legs on that venue freeze, unfilled orders are cancelled, and nothing new is sent until a fresh snapshot matches your ledger. Legs on the other venue finish only if the rulebook allows partial rebalances.

Why no sports contracts?

Sports event contracts face active state litigation and differing state rules. Ensemble starts with economics, crypto, weather and election markets and will revisit sports once the legal picture settles.