NEX7 — V1.3 PRED. PLATFORM / RESEARCH-GRADE
ESTABLISHED 2020 // MADISON,WI · CAMBRIDGE,MA

Biosignal Prediction Infrastructure

NEX7 helps researchers move beyond data collection — into forecasting episodes, personalized feedback, and real-world impact.

CH.01 · BIOSIGNAL STREAM
7d · ROLLING
SIGNAL RISK FORECAST
FS 256 Hz
[ 01 ] MISSION

The future of research isn't reactive — it's predictive.

Founded at the intersection of data science and clinical research infrastructure, NEX7 is building the tools that let researchers stop asking “what happened?” and start answering “what's about to happen, and what should we do about it?”

[ 02 ] WHAT WE DO

We build the connective tissue between biosignal data, predictive models, and the people generating that data.

/ 01

Predict episodes before they occur

Time-series models trained on multi-modal physiological streams forecast clinical events with calibrated confidence intervals — designed for the realities of free-living research.

/ 02

Empower study participants

Closed-loop interfaces return personalized insight to participants — turning passive data subjects into informed collaborators and improving adherence across longitudinal studies.

/ 03

Operate at research-grade fidelity

Validated pipelines, audit trails, and IRB-ready data handling. Built from day one for pharma and clinical partners who require provenance, not just dashboards.

[ 03 ] LEADING USE CASE

Migraine & headache research — one of the most data-rich, pattern-driven areas of chronic pain.

PRODROMAL WINDOW
T − 18h → EPISODE
SUBJECT 049
Validated context / Biocast

Surfacing the signals that precede an episode.

Biocast was initially developed and validated in the context of migraine and headache research. Our prediction engine is designed to surface the subtle prodromal signals that precede episodes — giving participants and researchers earlier, more actionable insight than any symptom diary alone can provide.

[ 04 ] HOW IT WORKS

Three stages, configurable end-to-end per study protocol.

01 / COLLECT

Capture what matters.

Participants log symptoms, behaviors, sleep, stress, and other relevant signals through a clean, accessible mobile experience. Researchers configure exactly what data matters for their study.

02 / PREDICT

Forecast what's coming.

Biocast's AI engine analyzes participant data to identify patterns and generate individualized episode risk forecasts — giving both researchers and participants a window into what's coming before it arrives.

03 / INTERVENE

Act on rising risk.

When risk levels rise, Biocast can deliver personalized, researcher-defined interventions — from medication reminders to behavioral nudges like rest, hydration, or light activity. Intervention logic is fully configurable per study protocol.

[ 05 ] FLAGSHIP

Biocast — the prediction layer for migraine & pain research, deployed across active studies today.

Biocast / v2

Forecast pain episodes hours before onset.

Biocast ingests wearable, EMA, and sleep signals; runs participant-specific models; and delivers calibrated risk forecasts to study teams and (optionally) participants. It replaces the post-hoc spreadsheet with a live, prospective clinical instrument.

6 hr
Median lead time
0.84
AUROC, validation cohort
SUBJECT 049 · 7-DAY VIEW FORECAST · LIVE
MONTUEWEDTHUFRISATSUN
OBSERVED PREDICTED RISK