CORPUS
Immediate action
102,361 8.2% · pacientes
acción →
Escalate · 24 h
Preventive medicine infrastructure
One layer of intelligence.Many conditions.Prioritized decisions.
Fragmented clinical data enters the Corpus AI Prioritization Engine and exits as action-ready cohorts, urgent, high-risk, chronic, or stable.
patients analyzed 1M+
infarctions prevented
sensitivity 93.8%
precision 95.8%
01 / the engine Prioritization Engine
The engine that turns data intoprioritized decisions.
A unified layer that reads any clinical data, computes risk with visible reasoning, and emits cohorts that are ready for action.
clinical chaos · fragmented data
Inputs
- HL7
- FHIR
- CSV
- EHR
- Labs
- Claims
- Notes
Corpus AI · Prioritization Engine
Cohorts
- Immediate action escalate · 24h
- High-risk intervention contact · 7d
- Active chronic management care plan
- Stable monitoring passive follow-up
prioritized cohorts
02 / population distribution Engine output · stratification
One population,four operational decisions.
The engine converts fragmented data into actionable groups. Each cohort arrives with its clinical action, no score to interpret, just execute.
n = 1,248,302 pacientesHigh-risk intervention
223,446 17.9% · pacientes
acción →
Contact · 7 d
Active chronic management
419,429 33.6% · pacientes
acción →
Care plan
Stable monitoring
503,066 40.3% · pacientes
acción →
Passive follow-up
Aggregate proportions across all clinical paths. 30-day window.
03 / the problem The problem
Today's system: fragmented, reactive, late.
Data in silos. Decisions made on incomplete information. Improvised prevention. No follow-up. 70% of healthcare cost stems from chronic disease that could have been prevented.
A unified system that reads any data, computes risk with visible reasoning, generates the plan, and keeps the loop active over time.
04 / relevance Signal over noise
What mattersnow, not what happened before.
Every patient accumulates years of signals. Corpus reads them together and continuously weighs which factor changes the clinical decision today.
Temporal context
An event's urgency depends on when it happened and how it relates to the rest.
Continuous re-weighting
The network recomputes with every new data point, without anyone retraining it.
Visible reasoning
Every active node appears as a modifiable factor in the final score.
- Abnormal ECG 3 days ago
- BP 164/102 9 days ago
- chest pain (reported) 22 days ago
- BP 158/96 85 days ago
- LDL 178 140 days ago
- BP 152/94 310 days ago
- LDL 165 540 days ago
- HbA1c 6.1 820 days ago
05 / how it works The loop
Five components.One single loop.
- 01 ingestion We connect any clinical data source HL7 · FHIR · CSV · EHR · PDF
- 02 standardization We turn data into clinical intelligence normalize · dedupe · map
- 03 risk + explainability We compute risk and explain why feature weights · modifiable factors
- 04 plan We generate the optimal intervention ranked by clinical cost-benefit
- 05 follow-up We follow the patient over time re-evaluated as risk changes
06 / architecture The model x-ray
↑ data flow Not a black box.An architecture you can inspect.
Three layers. Data flows up, from integration to compute to clinical action. Built for engineering teams, medical directors, and CTOs who need to understand, not just trust.
L3 · OUTPUT
Delivery & Clinical Action
Stratified lists, clinical reports, and dynamic dashboards that trigger clinical action, treatment adjustment, prevention programs, targeted intervention.
L2 · COMPUTE
Core Predictive Engine
Data reception and validation. The model processes the stream and feeds itself continuously, predictions update on their own when new data arrives.
L1 · INTEGRATION
Data Ingestion & Sync
Customer-system mapping, HIS, LIS, ERPs, via API, HL7, cloud buckets, SFTP, or manual CSV extraction as an immediate-impact pathway.
07 / team Leadership
The best talent,concentrated in five roles.
Five complementary roles, operations, technology, medicine, AI. Every clinical decision is signed by someone with medical practice; every technical decision, by someone who has held it up in production.
-
Juan Felipe Forero
Chief Executive Officer
-
Santiago Peláez
Chief Operating Officer
-
Juan Bernardo Benavides
Tech Lead
-
Juan Esteban Correa
Chief Medical Officer
-
José David Amorocho-Morales
Chief AI Officer
08 / clinical backing Supported by the best clinicians
Supported bythe best clinicians.
Latin American leaders in critical care, interventional cardiology, and public health who actively review the model, its clinical validity, and its implementation in the clinical workflow.
-
Dr. Guillermo Ortiz Critical Care · Intensive Care Medicine · Universidad El Bosque · Centro Policlínico del OlayaInternal-medicine physician, intensivist, and Latin American reference in critical care. His clinical review keeps the model’s operating bands meaningful for patients who pass through the ICU.
- Dr. Jorge Carrillo Cardiology · Interventional Cardiology · Interventional cardiologist
Interventional cardiologist with hands-on experience running the cath lab. Bridges the model’s prediction with precision-screening pathways (CCTA, CT-FFR, selective catheterization).
- Dr. Juan Esteban Correa Medicine · Palliative Care · Public Health · Chief Medical Officer · Corpus AI
MD, specialist in palliative care and public health. Owns the model’s clinical validity, the operating-band decisions, and the interface with international clinical guidelines.
Ready to see the platformon your data?
30 minutes with our clinical team. We look at your case. We respond in 2 business days.