The MDC agent workspace showing the unified patient conversation inbox.

An MDC patient conversation thread on mobile messaging.

The MDC patient context panel with medical history and insurance details.

Contextual MDC AI response suggestions offered inside a doctor's reply.

MDC attachment and image analysis within a patient conversation.

The MDC escalation view where a physician joins the conversation.

The MDC agent console on desktop with the conversation list and patient record.

The MDC patient-facing chat experience on a phone.

The MDC real-time urgency detection and triage board.

MDC

B2B2C, Healthtech, Healthcare, AI/ML

AI patient messaging: from day-long waits to 3-minute replies, 3x physician productivity.

AI-powered healthcare communication platform enabling 100+ trained human agents to represent 500+ physicians in routine patient interactions via WhatsApp/Telegram. Agents use AI-generated "digital personas" mirroring each doctor's communication style, with seamless escalation to the actual physician for complex cases.

platform

Web, PWA

verticals

B2B2C, Healthtech, Healthcare, AI/ML

team & duration

~6 FTEs, ~10 months

services

Digital product strategy, Digital product management, Digital product development

Challenge.

Physicians were overwhelmed by 15,000+ daily touchpoints, 60–70% of which were routine. Patients waited 1–2 days for basic answers, paid $150–300 per consultation for simple questions, and doctors burned out on repetitive inquiries instead of practicing complex medicine. The core tension: scaling doctor-patient communication without destroying trust, empathy, or clinical accuracy.

Solution.

An AI-powered agent workspace where 100+ human agents act as physician extensions using AI-generated "digital personas" that replicate each doctor's tone, analogies, and protocols. The system provides real-time urgency detection, contextual response suggestions in the doctor's voice, patient history at a glance, and transparent escalation ("Dr. Smith has joined the chat"). Patients always know they're speaking with a care team, not the doctor directly.

Impact.

Physician productivity

75–90 patients/day from 25–30

Time reduction

–50%

Cost per interaction

First-contact resolution

80%

without physician escalation

Quality scores

94%

Patient satisfaction

4.8/5

Physician burnout

–40%

No-shows

–60%

Patient retention

+30%

Patient capacity per physician

+45%

LTV: +30% patient retention and +45% capacity per physician directly increase lifetime value per doctor on the platform. Patients staying longer and engaging more frequently compounds revenue over time.

Retention: 30% increase in patient retention, 60% reduction in no-shows. On the B2B side, –40% physician burnout means lower doctor churn from the network.

Conversion: 3-minute response times vs. 1–2 day waits dramatically reduce patient drop-off. 89% of patients reported feeling "more heard" — higher engagement-to-consultation conversion. Patients previously avoiding care due to cost now engage at the $40 price point.

CAC: 50% cost reduction per interaction and new insurance reimbursements for digital interactions create a lower-cost acquisition and servicing channel.

UX: Agent-facing UX rigorously researched with 20+ medical professionals and messenger-pattern analysis. Patient-facing UX leverages existing WhatsApp/Telegram — zero learning curve. The persona system is a UX innovation giving patients continuity with "their doctor" even through an agent.

Features.

Unified agent inbox with smart prioritization (urgency + wait time)

AI digital persona system (communication fingerprint per doctor)

Real-time urgency detection and triage.

Contextual AI response suggestions in doctor's voice.

Patient profile with full medical history, insurance, family context.

Seamless doctor escalation with one-click join.

Multi-channel messaging (WhatsApp + Facebook + Telegram + Instagram)

Conversation search/filter by symptoms, meds, conditions, timeframes.

Quality assurance scoring and SLA tracking.

Auto-summarization of conversations.

Attachment/image analysis for preliminary assessment.

Drug interaction checker.

Subprojects.

Agent Workspace — core chat management, inbox, and productivity tool for 100+ agents.

Persona Engine — AI system for creating and maintaining doctor communication fingerprints.

Patient Context System — medical history, insurance, family data aggregation and display.

Escalation & Doctor Dashboard — real-time conversation monitoring, one-click physician join, post-conversation summaries.

QA & Compliance Module — conversation scoring, SLA tracking, HIPAA audit trails, physician review workflows.

Knowledge Management — dynamic FAQs, protocol library, drug interaction checker, insurance verification.

AI utilization.

Digital Persona System: Analyzes 100+ doctor communications to build a communication fingerprint per physician; generates response suggestions in each doctor's voice, tone, and preferred analogies.

Urgency Detection: NLP-based real-time triage of incoming patient messages, flagging cases that need physician review.

Medical Entity Recognition: Custom NER identifying symptoms, medications, conditions in patient messages.

Response Generation: GPT-4 fine-tuned on medical conversations provides contextual reply suggestions.

Attachment Analysis: Preliminary AI assessment of patient-uploaded images and documents.

Auto-Summarization: Condenses lengthy conversation histories so agents get quick context.

Pattern Detection: Identifies recurring patient concerns across conversation history.

Continuous Learning: Persona accuracy refines over time based on ongoing physician corrections.

This is a human-in-the-loop AI model — AI augments agents, agents deliver to patients, doctors oversee. Not autonomous AI healthcare.

Technologies and integrations.

Node

Nest

MongoDB

React

MUI

Firebase

OpenAI

respond.io

Datadog

Sentry

Vanta

Contentful