MedReach
An AI outreach system that fills medical webinars across voice and WhatsApp.
Each doctor receives a specialty-matched invitation in their preferred language, delivered by voice or WhatsApp according to prior response patterns. Consent, attendance and ROI are measured end-to-end across every campaign.
WHY IT WORKS
MedReach automates the outreach layer behind medical-education programmes — the telephonist operation that contacts doctors, delivers the invitation, handles objections and registers participants. The same workflow runs as a voice agent across multiple languages, with opt-outs and availability windows enforced at the system level and every interaction logged per doctor.
The core of the system is the data layer rather than the voice channel. A per-doctor availability model (PickupWindow) learns weekly patterns of reachability across campaigns. A specialty-matching layer (SpecialtyMatch) filters the outreach list against the clinical track of each webinar, so only clinically relevant doctors are contacted. Relevance and timing compound to increase acceptance rate and preserve long-term allow-list health.
Each contact within a campaign is scheduled dynamically. The cadence planner (CadenceConductor) selects the channel and timing of the next interaction based on the outcome of the previous one: a busy doctor receives a short WhatsApp message with the Zoom link; a declined call triggers a 72-hour silence period; a registered doctor receives a single reminder 60 minutes before the webinar. This logic protects the doctor's attention budget and keeps allow-lists healthy across campaigns.
The primary metric is attendance, since registration alone does not constitute a commercial outcome for the sponsor. AttendancePulse records live Zoom attendance per doctor; no-shows trigger an automated follow-up, attendees are routed to a short post-webinar NPS call. Every conversation opens with a logged disclosure; consent and opt-out records in ConsentLedger are enforced across voice, WhatsApp and email. Full per-doctor transcripts and audio recordings support pharma-grade audit requirements.
WHY MEDREACH
Outreach performance is determined by the data layer.
A manual call-centre applies a single script to the full roster within a shared working window, which imposes a structural ceiling on conversion. MedReach treats outreach as a data problem: voice delivery accounts for four services out of twenty in the architecture; the remaining sixteen govern targeting, timing, cadence, consent and attendance measurement.
The limitation of a manual call-centre is structural. A single script applied to the entire roster in a shared time window produces a predictable ceiling on conversion, regardless of how well the script is written. MedReach operates as infrastructure: each doctor is contacted inside a predicted availability window, on a topic matched to their specialty, with consent status and silence periods enforced by the system. Voice is four services out of twenty; the remaining sixteen constitute the data layer that produces the result.
- PICKUP WINDOW Each doctor is contacted inside an individually predicted availability window.
- SPECIALTY MATCH The outreach list for each webinar is filtered against the clinical track, so only relevant specialties are contacted.
- PEER SIGNAL Social-proof statements are generated from verified registration data, with per-institution and per-specialty counts drawn from the current campaign.
- SILENCE WINDOWS A decline triggers a configurable silence period (default 72 hours) across all channels, preserving long-term allow-list health.
Voice is four services out of twenty. The remaining sixteen constitute the data layer that determines campaign performance.
THE FLOW
How MedReach works — one campaign, six steps.
Each step is implemented as one or more microservices with a single responsibility. The full sequence runs in parallel, once per doctor, for every campaign.
-
Doctor profile
A single record per doctor: specialty, institution, preferred language, prior contact history and consent status. The roster is normalised on ingestion, so every downstream service reads the same canonical profile.
-
Targeting by clinical relevance
The webinar topic is mapped against each specialty profile. Only doctors for whom the topic is clinically relevant enter the outreach list; the rest remain untouched for this campaign.
-
Contact-time optimisation
A per-doctor model predicts the hours at which the doctor is reachable, based on prior interaction history. Each call is initiated inside that predicted window.
-
Multi-channel cadence
The channel and timing of every subsequent contact are selected individually based on the outcome of the previous one — voice call, WhatsApp message, or a defined pause after a decline. Each doctor follows an independent cadence.
-
Call, converse, register — in one session
An AI agent dials each doctor from a local number, speaks in their language with a voice tuned to their profile, answers the usual objections in real time, and — while still on the line — confirms the registration and sends the Zoom invite.
-
Attendance control and model calibration
A reminder is delivered 60 minutes before the webinar. Actual Zoom attendance is recorded per doctor. No-shows trigger an automated follow-up; attendees receive a short post-webinar NPS call. Pickup, conversion and attendance data from each campaign calibrate the model for the next one.
ARCHITECTURE · LAYERED BLUEPRINT
How the microservices wire up.
Read top to bottom. Channels are where people arrive, orchestration is how the AI thinks, domain is the startup-specific logic, data & identity keeps the source of truth, and platform is the invisible glue.
Scroll horizontally to see the full architecture.
INVENTORY
Every block in MedReach.
Each block is a microservice with a clear role — grouped by layer so you can see what belongs to channels, orchestration, domain, data and platform at a glance.
-
MEDICAL GRAPHDoctorGraphCanonical doctor profile
-
MATCHINGSpecialtyMatchTrack → eligible doctor list
-
TIMINGPickupWindowPer-doctor pickup ML model
-
SOCIAL PROOFPeerSignalTruthful peer-count strings
-
EVENTSWebinarDeskWebinar registry + waitlist
-
ANALYTICSAttendancePulseLive show-up tracker
-
COMPLIANCEConsentLedgerConsents + opt-outs
HOW TO SHIP IT
Quick-start checklist.
- 01 Import the doctor roster into DoctorGraph; enrich specialty, institution and language from the existing CRM.
- 02 Publish the first webinar in WebinarDesk; SpecialtyMatch auto-builds the eligible call list.
- 03 Turn on CadenceConductor; the first wave of calls runs inside each doctor's learned PickupWindow.
- 04 Wire AttendancePulse to the Zoom room; post-webinar NPS and no-show follow-ups trigger automatically.
WHAT'S NEXT
Compose MedReach in a visual workspace.
Log into AIM Engine 2.0, drop the blueprint on a canvas, fork the domain microservices you need, and deploy to your own domain. Everything shown here is already a real component.