Acoustic Crowd Intelligence · i2P LLC

AEGIS

Real-time acoustic prediction engine. Detects crowd eruptions 3+ seconds before they peak — in any stadium, any sport, any language.

App. 64/006,872 · Filed March 2026 NSF SBIR Phase I Applied Azure API · Live WKU Innovation Campus
5,700+
Validated Events
98.8%
Model Accuracy
500+
Minutes Processed
8+
Elite Competitions

How AEGIS works.

No cameras. No player tracking. No broadcast feed required. AEGIS reads the crowd's acoustic energy as a living signal — and fires before the event peaks.

  1. Step 01 — Capture
    Listen
    30-second audio windows from the live venue feed. Any stadium. Any sport. Any language.
  2. Step 02 — Extract
    Analyze
    6-point acoustic feature vector: amplitude envelope, rise/fall time, spectral centroid, RMS energy, duration, zero-crossing rate.
  3. Step 03 — Classify
    Detect
    Adaptive spike detection + Random Forest classifier. QAOA validation on borderline events. Distinguishes eruptions, swells, groans, and whistles.
  4. Step 04 — Alert
    Fire
    Anticipatory alert fires 3+ seconds before event peak. REST API response. Haptic trigger. Broadcast overlay. <100ms latency.

The data speaks.

Validated against real match broadcast audio across World Cup, Champions League, Premier League, and La Liga. 5,700+ labeled crowd acoustic events and counting.

AEGIS validation results by match across elite competitions
Match Competition Duration Validated Events Accuracy
Croatia vs BrazilFIFA World Cup QF55 min95190.3%
Barcelona vs AtléticoLa Liga54 min41100%
Real Madrid vs BayernUCL Semi-Final100 min19497.9%
PSG vs Inter MilanUCL Final55 min126100%
Argentina vs CroatiaFIFA World Cup SF56 min88100%

Three lanes.

One acoustic intelligence engine powering broadcast, fan experience, and accessibility. API-first. No new stadium hardware required.

● Live
AEGIS Azure API
REST API processing crowd audio and returning validated spike data in real time. Deployed on Azure Container Apps. <100ms latency. Per-event or annual enterprise license.
aegis-api.wittyground-7054d6c9.eastus.azurecontainerapps.io
● Ready
Broadcast Intelligence SDK
White-label SDK for leagues, OTT platforms, and second-screen apps. Embeds AEGIS directly into existing broadcast pipelines. No hardware. Tier 1–3 enterprise licensing.
Enterprise licensing — contact for pricing
In Prep
Battito™ Haptic Wristband
Consumer haptic device receiving AEGIS alerts in real time. Italian: "heartbeat." Dual audience: neurodivergent fans (sensory safety) + Kings League & esports fans (hype immersion). Patent application in preparation. Via i2P Haptic Designs.
i2P Haptic Designs · Target launch Phase II

Protected on four fronts.

Patent, dataset, federal grant pipeline, and pipeline lock-in. Each layer reinforces the others.

Patent Filed
⚖️
AEGIS Core Method
Provisional App. 64/006,872 filed March 16, 2026. Non-provisional deadline March 16, 2027. Claims cover: anticipatory 3-second lead, 6-point feature vector, adaptive percentile baseline, multi-sport dependent claims.
Data Moat
🏋️
Training Dataset
5,700+ labeled acoustic crowd events from World Cup, UCL, Premier League, and La Liga. Proprietary. Competitors starting from scratch would need 12–18 months to replicate comparable signal coverage.
Non-Dilutive
🔬
NSF SBIR Phase I
$275K Phase I applied. NSF SBIR provides non-dilutive capital validation and positions AEGIS as a research-grade innovation — critical for government, stadium, and university partnerships.
App. 64/006,872
Provisional patent filed March 16, 2026 · Non-provisional deadline March 16, 2027
Pro bono patent counsel program (pending assignment) · i2P LLC · Bowling Green, Kentucky

Let's talk.

Founder & CEO
KaTerrell Kennedy
Founder & CEO · Principal Investigator
General & Investment Inquiries
Location
WKU Innovation Campus · Bowling Green, KY
Company
i2P LLC · International Institute for Integrated Performance