AITHE CONNECTOR0
 
Initialising 0%
Remote team · UK, Gulf and US East hours UK · GULF · US EAST | 14 platforms shipped.NET · Angular · Python · NodeOpenAI · n8n · Twilio · Azure14 platforms shipped
ClarvoTechAI
Book a call
AI & Automation
Build & Run
Try it

Ask us anything

Message us on WhatsApp. A founder replies, works out what you need, and offers a call if it makes sense.

Open WhatsApp
↑↓ navigate↵ openesc close
Home / Work / Unified inbox
Healthcare · 60 staff · Multi-channel AI

Four inboxes became one thread.

AI Integration·Systems Integration·5 months·figures illustrative until client sign-off
Channels unified
0
Typical reply time
0s
Sorted without a person
0%
Passed over at the right time
0%

The problem

Patients got in touch four different ways and each one landed somewhere else. Phone calls on one system, texts on a shared mobile, WhatsApp on whoever happened to be holding the work phone, and website enquiries in an inbox nobody owned. Nobody could answer the one question that mattered: has this person already been replied to?

The failure mode was never a missed message. It was two people answering the same person differently on two channels.

What we built

One place where everything arrives. No matter how someone gets in touch, it joins the same conversation with that person, in order. On top of that, AI writes a first draft of the reply, shows the staff which record it used, and hands over to a person the moment it is unsure.

  • Replies go back the same way they came in, so nobody gets a text back to a phone call
  • Photos and documents stay attached, whichever app they arrived through
  • Reception and clinical staff see different things, because they should
  • We tested when it should hand over to a person, rather than guessing

Why it is on this page

Because it is the same approach we use for our own enquiries: retrieval over what we actually do, plus clear hand-over rules to a person. Message us and see.

OpenAITwilioWhatsApp Cloud APIwebhook ingestionvector retrievalRBAC.NET