L1 Context Ingestion L2 Symptom Structuring L3 Differential Generation L4 Treatment Pathway L5 Follow-up Protocol
1docAI Clinical OS CRE
CLINICAL REASONING

CRE

The Consultation Reasoning Engine. A 5-layer AI framework that structures every 1doc patient consultation — from first contact through diagnosis, treatment planning, and follow-up.
Mission Zero is achieved not through episodic treatment,
but through continuous visibility, continuous intervention, and continuous accountability.

ARCHITECTURE
5-Layer Framework

INTEGRATION
Evidence Graph + DEC

AUTHOR
Dr Eugene Loke, CMO, 1docAI

What is the Consultation Reasoning Engine?

The Consultation Reasoning Engine (CRE) is 1docAI's AI framework for structuring clinical consultations. Every time a 1doc patient interacts with a family physician, health coach, or the AVA virtual assistant, the CRE operates in the background — synthesising the patient's Evidence Graph data, the presenting complaint, and clinical guidelines to support structured, consistent, high-quality care.

The CRE is not a diagnostic tool that replaces the physician. It is a reasoning scaffold that ensures every consultation follows a rigorous clinical pathway — reducing variation, surfacing relevant history, and flagging considerations the clinician needs to address.

The 5-Layer Architecture

One of the fundamental challenges in community healthcare is variation — the quality and thoroughness of a consultation should not depend on which physician a patient sees, how busy the clinic is, or how recently the doctor reviewed the relevant guidelines. The CRE addresses this by providing a consistent reasoning scaffold across all 1doc clinicians.

This is particularly important for 1docAI's ambition to operate across 50 AI clinics and manage 300,000 patients under continuous care. At that scale, clinical consistency is not just a quality metric — it is a patient safety imperative.

CRE and clinical consistency at scale

The CRE operates across five sequential layers, each building on the outputs of the previous:

L1
Context Ingestion
Pulls the patient's relevant Evidence Graph data — recent vitals, chronic conditions, current medications, previous consultations — and frames the clinical context for the session.
L2
Symptom Structuring
Organises the presenting complaint using standardised clinical frameworks, surfacing differentials and flagging any presenting features that require immediate DEC evaluation.
L3
Differential Generation
Generates a ranked differential diagnosis list weighted against the patient's individual risk profile from the Evidence Graph — not generic population statistics.
L4
Treatment Pathway Mapping
Maps evidence-based treatment pathways against the patient's current medications, known allergies, and lifestyle context — flagging contraindications before prescription.
L5
Follow-up Protocol
Generates a post-consultation monitoring protocol — specifying what DEC should watch for, when the next touchpoint should occur, and what outcomes should trigger escalation.

Frequently Asked Questions

CRE Architecture
Layers 5 sequential
Evidence Graph integration L1 + L3
DEC integration L2 + L5