Before
- 100 to 200 people on phones
- Call cabins and phone lines to run
- Notes typed after every call
An AI that calls patients, so staff don't have to.
Care teams set it up once. It then calls every patient on schedule, asks simple questions like "Did you take your medicine today?", and shows staff who needs help.
2.5M+
Calls made
14+
Healthcare institutions onboarded
1,500+
Hours saved every month
At a glance
- Client's vision
- Cut manual calling and keep a regular, easy check on every patient
- Industry
- Healthcare
- Services
- We joined the development team of an established healthcare business, automated their manual work with AI, and built an AI voice app.
- Engagement
- Retainer
- Live product
- Get live product link
Client testimonial
“Working with Hari and his team has been a fantastic experience. They’re responsive, thorough, easy to communicate with, and genuinely great people to work with. I’ve really enjoyed collaborating with the team and getting to know them along the way. They’ve been a reliable partner throughout the project, and I’d be happy to work with them again.”
Monty Alexandar
CEO★★★★★
What was the problem?
Checking on thousands of patients used to need call centre staff phoning everyone, one by one.
After
- Set up a round of calls once
- AI calls every patient
- Summaries and urgent cases land on one screen
How we solved it
Making an AI talk was the easy part. The care team needed every call to end with a clear next step. So we built everything around that call.
Every call moves through three steps
- 1
Connect
- Set up a campaign: one round of calls to a list of people
- Choose who to call, when and why
- The platform places and tracks every call
- 2
Understand
- Each call is recorded and turned into text
- AI writes a short summary and a clear result
- Nobody has to listen to recordings
- 3
Act
- The result starts the next step by itself
- Anyone who asks for help goes straight to an operator
- The team sees every call and result on one screen
The hard parts
Handling lots of calls
How do you make thousands of calls without slowing everything down?
We keep the heavy jobs, like storing recordings and analysing calls, separate from the app people use. So the app stays fast, however many calls are made.
Understanding every call
How do you turn a phone conversation into something a team can use?
Every call is turned into text. AI then reads it and writes a short summary and a clear result, so staff know what happened without listening to recordings.
Getting better over time
How do you keep the AI improving after launch?
Staff can flag any call in one click. That feedback is used to improve the AI and the product, again and again.
Features we built
01Custom voice agentWhen they sign up, each care provider builds its own AI caller by choosing who it speaks to, how it sounds and the tone it uses.
02Unlimited campaignsCreate a campaign for any need, such as a daily call asking patients if they took their medicine: press 1 for yes, 0 for no.
03Easy contact importAdd the people to call by uploading a spreadsheet, typing them in, or connecting their existing customer list.
04Call summariesOne dashboard with success rates, call history, transcripts and recordings, with filters to find any call.
05Call schedulingPick the dates and times, weekdays or weekends. Redialing settings for unanswered calls.
06Support ticketsRaise a ticket on any call. It goes to the right support team automatically, with email updates and a board to track progress.
07Performance reportsWeekly and monthly reports show how every campaign is performing.
08Admin dashboardOne place to update voice agents, create campaigns, invite users, set who can access what, and see every call and ticket.
09Subscriptions and billingMonthly plans or pay-as-you-go, with payments handled through Stripe.
01 / 09
Where it is today
It started as an AI that could hold a phone call. Today it is a full platform behind millions of calls.
Calls made
Staff cost saved every month
hours saved every month
Healthcare institutions actively using
Tools we used
Built with
Bubble
JavaScript
Claude
Lovable
OpenAI
AI models
Gemini
Multimodal AI
n8n
Workflow automation
AWS
Cloud infrastructure
Figma
Product design
Railway
Application hosting
React
User interfaces
Python
Automation & AI
TypeScript
Typed applications
Node.js
Server applications
Integrated with
Retell AI
AI voice
Slack
Notifications
SendGrid
Notifications
GoHighLevel
Customer records
Freshsales
Customer records
Salesforce
Customer records
Pipedrive
Customer records
Make
Connects the steps
CSVbox
Uploading lists
Sign-in
Persona
ID checks
Docusign
Signing agreements
Typesense
Search
Stripe
Payments
Let's Build Together
Stop wasting time on manual processes. Start building a self-running business.