Find alpha
through event risk.

Untitled watches prediction markets for large probability moves. When one lands, the terminal lays out how assets have historically responded to similar shocks, with sample sizes and intervals attached, and stresses those histories against your portfolio weights. Nothing in it can place an order.

Open the terminal See how it works

01 DETECT

A probability shock enters the scanner.

When a prediction-market probability moves sharply, the scanner picks it up and ranks it by attention. In this snapshot, "Will Bitcoin dip to $63,000 on July 27?" fell 51.9 percentage points in an hour, a move in the 98.2th percentile of comparable history.

The scanner stays keyboard-first. Filter by time, category, venue, linked asset, research status and magnitude, pin a watchlist, and share any view as a URL.

MOVE: -51.9 PP / 1H
UNUSUALNESS: P98.2 · n=84
MARKET QUALITY: B
ATTENTION: 68 / 100

UNTITLEDMONITOR

EVENT SCANNERATTENTION-RANKED

UTC	EVENT	ΔPP	P0	P1	STATUS
03:50	Will Bitcoin dip to $63,000 on July 27?	−51.9	52.4%	0.5%	WATCH
03:00	WTI Crude Oil (WTI) Up or Down on July 28?	−14.5	47.0%	32.5%	RADAR
04:10	Will WTI Crude Oil (WTI) hit (LOW) $80 in July?	−13.0	84.5%	71.5%	RADAR
03:40	Will “Spider-Man: Brand New Day” score at least 95 on Rotten Tomatoes?	+13.0	10.5%	23.5%	RADAR
03:30	Will the Fed decide differently in the next three decisions?	+10.0	62.5%	72.5%	WATCH
03:20	Will the Fed Pause–Pause–Pause in the next three decisions?	−10.0	37.5%	27.5%	WATCH
04:30	Will Elon Musk post 260-279 tweets July 21 to July 28?	+7.5	12.0%	19.5%	RADAR
03:09	Will WTI Crude Oil (WTI) hit (LOW) $75 in July?	+7.0	15.0%	22.0%	RADAR

02 MAP

Historical response arrives with its evidence attached.

Each reviewed asset channel reports what happened across comparable past shocks: the median response at five-minute, fifteen-minute and one-hour horizons, the 95% interval around it, the sample count behind it, and the named chronology, leakage, out-of-sample and multiple-testing checks it passed or failed.

Uncertainty stays on screen. Wide intervals print wide, and every relationship carries its review status.

HORIZONS: 5M · 15M · 1H
INTERVALS: 95% · BOOTSTRAPPED
VALIDATION: 19 NAMED CHECKS
PATH ANALYTICS: VELOCITY · DRAWDOWN · VARIATION

EVENTWill Bitcoin dip to $63,000 on July 27?

ASSET	MED BPS	95% INTERVAL	N	STATUS
SOL	−2.0	−22.7 / +2.1	66	WATCH
ETH	+7.0	−7.6 / +12.5	66	WATCH
BTC	+3.5	−12.7 / +5.2	66	REJECTED

03 STRESS

The event meets your book privately.

Manual BTC, ETH and SOL weights stay in the browser session. A connected Alpaca paper account arrives read-only and pre-aggregated into coarse exposure bands, coverage and concentration.

The stress ladder and weighted historical stress path scale each reviewed channel across your weights, with browser-local limits and alerts. Try it: the ladder on the right recomputes as you type.

MANUAL WEIGHTS: BROWSER-LOCAL
CONNECTED RISK: COARSE BANDS ONLY
STRESS PATH: WEIGHTED · MAX DRAWDOWN
ORDER PATH: NONE

PORTFOLIO

BTC%ETH%SOL%

GROSS80%

HORIZON	WEIGHTED HIST MEDIAN	PER $1M
5M	−0.5 BPS	−$51
15M	+0.2 BPS	+$19
1H	+2.9 BPS	+$285

Historical medians from the demo event's reviewed channels, scaled across your weights.

04 QUANT

A working options and factor workbench.

Compose up to twelve session-local option legs, solve implied volatility, aggregate first and second-order Greeks, chart the position value curve with computed breakevens, and reprice the full position across spot, volatility and time scenarios.

Public-factor VaR and expected shortfall come from bounded public covariance. The curve on the left is the live model: switch implied volatility and read it with your pointer.

GREEKS: FIRST + SECOND ORDER
VALUE CURVE: BREAKEVENS · CROSSHAIR
SCENARIOS: SPOT × IV × TIME
FACTOR RISK: VAR · ES · CORRELATION

QUANTBTC 30D ATM CALL · P(ITM) 47%

Solid: model value today. Dashed: value at expiry. Premium $4,906 per contract.

One API. Every venue.

The snapshot, responses and stress model as clean Python.

untitled / api / python sdk

# pip install untitled
from untitled import Untitled
 
client = Untitled(api_key=KEY)
 
# Highest-attention event right now
event = client.events.top()
print(event.title, event.delta_pp, "pp")
 
# Or scan the universe with scanner-grade filters
for e in client.events.list(
    category="rates_macro",
    venue="polymarket",
    min_move_pp=10,
    limit=20,
):
    print(e.utc, e.title, e.attention)

PolymarketLive

KalshiLive

PredictItNot yet

LimitlessNot yet

Research at scale.

One snapshot carries 1,454 monitored events and 1,430 probability shocks from the last 24 hours, feeding a registry that promotes and demotes asset relationships through named statistical checks.

EVENTS MONITORED: 1,454
SHOCKS · 24H: 1,430
UNDER WATCH: 162
CANDIDATE: 22
REJECTED: 32

Built read-only, end to end.

ACCESS: Server-enforced sessions · signed HMAC cookies · 8-hour expiry
DATA BOUNDARY: One fixed research snapshot source, validated server-side: schema, size, event caps, and a recursive privacy filter that rejects account, balance, position, order, wallet, and credential material
MARKET DATA: Bounded public Coinbase statistics and hourly candles: read-only, capped, and time-limited
BROKER SCOPE: Alpaca paper OAuth only, read-only by construction; write and trading scopes are refused at the callback
EXECUTION: No order ticket, no execute button, no account-write path anywhere in the product
ANALYTICS: The terminal loads no analytics, no remote scripts, no remote fonts

Find alpha with Untitled.

Access is for authorized users. The terminal opens on the live research feed with nothing to install.

Open the terminal