We read the story in your glucose long before a diagnosis does, so problems can be caught and prevented while there is still time to change them. Clinically validated, and built with clinicians.
A single average (HbA1c, or mean glucose) collapses a living signal into one number, and hides the volatility that actually damages tissue. That kind of abnormal glucose behaviour, before it meets the threshold for diabetes, is called dysglycemia. Bio-Conscious models the whole curve: its shape, timing, variability and trajectory. We cluster patients into distinct glucotypes, the metabolic types defined by the shape of a person's glucose curve, then forecast where each curve is heading. That is the shift the field has been waiting for, from measuring the past to modeling the future.
Same number on the chart. Opposite risk in the body. The average is what every clinic sees. The shape is what we model.
Bio-Conscious clusters patients into four glucotypes by curve shape: stable responders, post-meal spikers, variable/brittle, and sustained-elevated, separating patients who share the same average glucose into distinct metabolic phenotypes with different risk and intervention.
There is no single “diabetes.” By clustering curve shapes into glucotypes, our models separate patients who look identical on paper into distinct metabolic phenotypes, each with its own risk profile and its own intervention. This is the engine behind moving care from reactive to predictive to preventive.
Endobits is clinical decision support for continuous glucose monitoring (CGM). It reads glucose data from Dexcom, Abbott and Senseonics sensors, flags the patients most at risk, and turns that work into billable remote patient monitoring under Medicare (CPT 99453, 99454, 99457, 99458). Clinics earn revenue today for the preventive care they are already positioned to deliver, and build toward prevention at scale.
Endobits forecasts glucose up to 12 hours ahead of the live CGM reading, catching predicted hypo- and hyperglycemic events before they happen, while other platforms stop at the current value.
Our defensibility is a whole stack, not a single model. Each level builds on a proprietary understanding of glucose metabolism, compounding into an early-detection engine that extends far beyond diabetes.
Validation of the Endobits prediction engine: the forecast line runs 12 hours ahead of the live CGM signal, and the subsequently observed glucose follows the predicted trajectory.
For any CGM platform, the cost of acquiring a patient dwarfs the cost of keeping one, and the patients who churn are the ones who stop seeing value in their data. Endobits turns raw readings into something a patient acts on every day, which is what keeps them on-sensor. The math at platform scale is not subtle.
Illustrative model. Adjust the inputs. Annual revenue retained = base × revenue/patient × retention gain. Scenario inputs, not a forecast.
CGM adoption is exploding, but the data dies in dashboards. Healthcare still reacts to metabolic disease instead of preventing it. The unused signal inside glucose data is one of the largest untapped opportunities in metabolic health.
The model is the easy part. The moat is everything around it, and most of it takes years and clinical relationships to assemble. We already have.
A generic foundation model can predict the curve. Only forward clinical-outcome labels tied to the glucose signal tell you what that curve meant for the patient, and that labeled, longitudinal dataset takes years of clinical partnership to build, not a quarter of engineering.
Not a research notebook. A prediction engine hardened inside a live clinical product against real-world CGM noise, gaps and sensor switches, the unglamorous work that separates a demo from a deployment.
Proprietary clustering of curve shapes into distinct metabolic phenotypes, the layer that turns a forecast into personalization, with early longevity indications competitors don't have.
We integrate Dexcom, Abbott and Senseonics, including the only implantable CGM. A single device-maker building this in-house builds a walled garden; we are the neutral intelligence layer that spans all of them.
Nearly a decade of R&D and research presented at ADA's Scientific Sessions, born from a study at BC Children's Hospital. In healthcare, trust is earned slowly, and it can't be cloned in a sprint.
Endobits is HIPAA-compliant and runs on top of FDA-cleared sensors from Dexcom, Abbott and Senseonics. It is designed to operate within the established Medicare remote-patient-monitoring and chronic-care-management framework, so the clinical value it creates is billable today, not contingent on a new reimbursement category.
Trusted by the ecosystem we build on
As part of a $17.3M PacifiCan investment across eight B.C. technology companies, Bio-Conscious Technologies received $1.5 million to commercialize Endobits, its AI platform that detects medical events at their earliest stage for patients with diabetes, for hospitals across Canada and the United States.
Read the announcement →The thesis
Glucose is the most-tracked, least-used signal in medicine. We built the half that acts on it.
Dysregulated glucose metabolism drives the core mechanisms of aging itself, from cellular senescence to mitochondrial decline. That makes the 5th vital sign a lever not just for disease, but for healthspan. We built an interactive guide to show exactly how the connections map.
Insulin/IGF-1, mTOR, AMPK and the sirtuins are the nutrient-sensing machinery. Every excursion is a direct instruction to them, and chronic hyperglycemia leaves the switches jammed in growth mode.
Endobits lens The hallmark our data reads most directly. This is the core signal Endobits models.
It isn’t the average that wears mitochondria down, it’s the swing. Each rapid excursion floods them with reactive oxygen species, and the repair debt compounds decade over decade.
Endobits lens Curve volatility, not mean glucose, is our earliest non-invasive read on mitochondrial strain.
Sustained hyperglycemia and advanced glycation end-products (AGEs) drive cells into senescence — and senescent cells handle glucose worse, so the loop tightens with every year.
Endobits lens Excursion frequency tracked over months is our proxy for accumulating senescent load.
High glucose switches on NF-κB and the cytokine cascade behind “inflammaging” — the common thread tying metabolism to nearly every age-related disease.
Endobits lens Glycemic instability tracks inflammation more tightly than A1c does.
Glucose flux sets the acetyl-CoA and NAD+ pools that write DNA methylation and histone marks. This is metabolic memory: why a bad glycemic decade still shows years after control returns.
Endobits lens Glucotype shape, not level, is where we look for the direction of epigenetic drift.
Glucose glycates and cross-links proteins faster than the chaperone and clearance systems can keep up. A1c is that reaction measured on hemoglobin — but it happens everywhere.
Endobits lens We integrate area above range over time: the cumulative glycation dose, not a 90-day average.
mTOR suppresses autophagy whenever glucose is high. The troughs between meals and overnight are the only windows the cell gets to clean house — and most people have fewer than they think.
Endobits lens We measure time-in-trough: the hours your autophagy is actually permitted to run.
Hyperglycemia-driven oxidative stress damages DNA and, in the same stroke, impairs the repair pathways meant to fix it.
Endobits lens High-variability signatures are our marker for oxidative DNA load.
A dysglycemic environment blunts progenitor-cell function and the tissue repair that depends on it — visible in how slowly a diabetic wound closes.
Endobits lens Years of cumulative exposure, which only CGM captures, is the upstream marker of regenerative decline.
AGE–RAGE signaling and a shifted adipokine balance distort the chemical messaging between tissues, turning a metabolic problem into a whole-body one.
Endobits lens Glycemic state is a systemic input to that network — and one of the few you can actually change.
The oxidative load from repeated excursions is associated with faster telomere shortening; diabetes consistently tracks with shorter telomeres.
Endobits lens A longitudinal question our dataset is built to answer at scale.
Diet and glucose reshape the gut microbiome, which feeds straight back on glucose metabolism — a bidirectional loop only now being mapped.
Endobits lens CGM is the highest-resolution window we have into that loop.
hallmarks of aging with a documented link to glucose metabolism. The most-tracked signal in medicine is also one of the most upstream.
The opportunity is a timing window, not a better app. Three independent shifts are converging, and the intelligence layer that sits at their intersection captures all three.
Continuous glucose sensing is moving beyond diabetes into prevention, performance and consumer health, and the volume of data is compounding.
Production-grade machine learning makes real-time, personalized forecasting reliable enough for clinical use, not just research.
Health systems and payers are moving from reactive treatment toward prevention and remote monitoring, with reimbursement following.
“Integrating CGM data with AI like Endobits is a significant advancement in diabetes care, giving clinicians actionable insight in real time.”
Glucose is the most-tracked, least-used signal in medicine. We've spent nine years turning it into a defensible, revenue-generating engine for disease prevention and longevity. And we're just reaching Level 3.
Email us, contact@bioconscious.tech
Moms+ studies how glucose patterns in pregnancy shape a child's long-term metabolic health. Leave your email and we'll notify you the moment enrollment opens.
A continuous glucose monitor, a small sensor worn on your arm or abdomen.