Skip to content
Quantitative Intelligence

Daily intelligence to monitor.

Composite z-score with per-signal Bayesian attribution. Point-in-time provenance for every number. Three-layer anomaly consensus.

25 companies monitored
US LARGE-CAP
21 data sources
PUBLIC FEEDS
13 quantitative models
PEER-REVIEWED
06:00 UTC daily brief
WEEKDAYS
25 companies monitored
US LARGE-CAP
21 data sources
PUBLIC FEEDS
13 quantitative models
PEER-REVIEWED
06:00 UTC daily brief
WEEKDAYS
Capabilities

13 quantitative models.
Zero black boxes.

01
Composite Score

8-signal z-score aggregation with per-signal Bayesian conviction weighting. Equal-weight outperforms optimized per research.

02
Anomaly Ensemble

Mod-Z + ECOD + CUSUM + COPOD with 2/4 consensus voting. Dramatically reduces false positives vs single-detector approaches.

03
Regime Detection

Gaussian HMM 3-state volatility classification with Kalman-filtered transitions. Calm, transitional, and stressed regimes.

calm transitional stressed
04
Kalman Health

Latent state estimation with confidence bands. A widening confidence interval signals degraded data coverage, not a change in health.

95% CI
Process

Three steps.
Every weekday morning.

config.yaml
Data Pipeline

21 sources.
Zero manual input.

Every signal is automatically collected, validated, and versioned. Point-in-time provenance for every number.

21
Data sources
06:00
UTC daily
100%
Point-in-time
Data Sources All operational
SEC EDGAR
Regulatory
live
Form 4 Insiders
Insider
live
Stock OHLCV
Market
live
Options IV
Derivatives
live
GDELT BigQuery
News
live
Bluesky Jetstream
Social
live
Reddit PRAW
Social
live
Polymarket
Prediction
live
FRED Macro
Economic
live
+ 12 more sources
Platform

Numbers that
speak for themselves.

Live
0
Companies monitored
0
Public data sources
0
Quantitative models
06:00
UTC daily brief
Data Sources

21 public feeds.
Every signal traced to its origin.

SEC EDGAR
Form 4
13D/13G
13F
XBRL
Stock OHLCV
Options IV
Forward Markets
GDELT BigQuery
GDELT DOC
Bluesky
Reddit
StockTwits
Google Trends
Wikipedia
Adzuna
openFDA
EPA ECHO
FINRA
Polymarket
FRED
SEC EDGAR
Form 4
13D/13G
13F
XBRL
Stock OHLCV
Options IV
Forward Markets
GDELT BigQuery
GDELT DOC
Bluesky
Reddit
StockTwits
Google Trends
Wikipedia
Adzuna
openFDA
EPA ECHO
FINRA
Polymarket
FRED
FRED
Polymarket
FINRA
EPA ECHO
openFDA
Adzuna
Wikipedia
Google Trends
StockTwits
Reddit
Bluesky
GDELT DOC
GDELT BigQuery
Forward Markets
Options IV
Stock OHLCV
XBRL
13F
13D/13G
Form 4
SEC EDGAR
FRED
Polymarket
FINRA
EPA ECHO
openFDA
Adzuna
Wikipedia
Google Trends
StockTwits
Reddit
Bluesky
GDELT DOC
GDELT BigQuery
Forward Markets
Options IV
Stock OHLCV
XBRL
13F
13D/13G
Form 4
SEC EDGAR
Methodology

Academic foundations.
Peer-reviewed models.

Every model in Sentinel is grounded in published research. No proprietary black boxes.

Bayesian ECOD HMM Kalman FinBERT Dempster-Shafer PELT Granger
Anomaly Detection

Li et al., "ECOD" (IEEE TKDE 2022). Adams & MacKay, "Bayesian Online Changepoint Detection" (2007). Modified Z-score with MAD.

Regime Detection

Hamilton (1989) on regime-switching. Nystrup et al., "HMM for Financial Regime Detection." Kalman (1960).

Insider Trading

Cohen, Malloy & Pomorski, "Decoding Inside Information" (JF 2012). Opportunistic vs routine classification.

Linguistic Analysis

Cohen, Malloy & Nguyen, "Lazy Prices" (JF 2020). Loughran-McDonald sentiment. Araci, "FinBERT" (2019).

Pricing

Simple, transparent
pricing.

Start free and scale as you grow. No hidden fees, no surprises.

Monthly Annual Save 17%
01

Research

For individual researchers

£0 /month
  • Up to 10 companies
  • Daily email brief
  • Web report access
  • 13 quantitative models
  • 21 data sources
  • No API access
  • No webhook alerts
Start Free Trial
03

Enterprise

For institutional teams

Custom
  • Everything in Pro
  • Custom company universe
  • Dedicated support
  • SLA guarantee
  • On-premise option
  • Custom integrations
Contact Us

All plans include 14-day trial, weekday briefs, and point-in-time provenance. Compare all features