Endobits forecasts near-term glucose events and turns them into a priority list: who's heading for trouble, and when. We surface the signal so care reaches the right patient before a complication does. The clinical call stays yours.
Overnight
Severe nocturnal events
Our sharpest window is sleep — when severe lows go unwitnessed. Endobits forecasts overnight hypo- and hyperglycemia 1 to 9 hours out, above 90% accuracy, so the team acts before morning.
Optimal window · during sleep
~95%+
Forecast accuracy, real patients
Accurate means the forecast lands within 10% of the real glucose value — scored across real-world and observational studies.
Real-world study · 142 participants
400+ patients
Tested, incl. public benchmark
Validated across multiple studies and on OhioT1DM — the public benchmark diabetes-AI researchers measure against.
Multi-study · OhioT1DM benchmark
Sharpest overnight. Viable to 13 hours.
Prediction accuracy by forecast horizon · “accurate” = within 10% of the real glucose value
Optimal — nocturnal event window (~1–9 h, during sleep)
Algorithm viability — uncertainty widens
90% accuracy floor · error is tightest (± 1.7%) overnight and widens the further out we forecast Other methods lose accuracy past ~30 minutes — off the left edge of this chart
Hours 1–9 are where it counts — the overnight window, tightest predictions, severe events flagged during sleep. Past that, the model holds out to 13 hours with widening uncertainty: headroom, not a guarantee.
In live clinical use · US
Prediction is the start. This is what changes.
43%
Fewer severe hypos
With early warnings in hand, care teams cut severe hypoglycemic events sharply — the events that put patients in the ER.
Live clinical patients · US
+6 hrs
More time in range, daily
On average, patients gained about six more hours a day in the healthy glucose range — the metric tied to better long-term outcomes.
Live clinical patients · US
250%
Higher per-patient income
Endobits opens reimbursable remote-monitoring workflows — a 250% lift in per-patient income from care you're already positioned to deliver.
US reimbursement · RPM / CCM
“We can only scratch one back at a time, but we can teach many patients together — and each is likely to teach another.”
Dr. Elliott P. Joslin · founder, Joslin Diabetes Center
Give your CDCES team superpowers
Not a clinician replaced by an algorithm — or the other way round. A third capability: your specialists' judgment, with the reach to apply it to every patient at once. Endobits triages the full 1,500-patient panel in seconds so the team's hours land where they change outcomes. That reach is what drives the 250% per-patient income.
CDCES team + Endobits
~ 1.5 seconds
CDCES team alone
~ 5.5 months
We surface the signal. The call stays yours.
What this is — and isn't. Endobits is clinical decision-support software used under a clinician's direction. It is not a medical device, not a diagnosis, and not a substitute for professional medical advice. Prediction figures come from internal, observational, real-world, and benchmark testing; clinical-outcome and revenue figures come from live and observational use in US clinics. Individual patient and clinic results vary and are not guaranteed; reimbursement depends on payer, coding, and documentation.