Case Studies | Healthcare & Life Sciences

A production-scale AWS foundation for a multi-channel medical assistant

Multi-Channel Platform Build

About

A Healthcare & Life Sciences business working with Cloud Combinator on AWS. The client is anonymised at their request.

Challenge

The build had three focus areas, each of which had to hold at production scale before the platform could be handed over.

Running a multi-channel assistant at production scale

The platform serves many clinics at once across three channels, so it had to scale per tenant and per workload and run each part independently. That meant separating the network edge, the compute, the AI and voice services, and the data, so no single component becomes a bottleneck under load.

Moving to AWS-native AI and voice without disruption

The assistant relied on external providers for its language model and speech. Introducing Amazon Bedrock, Transcribe and

Keeping tenants isolated and the platform secure

With patient conversations flowing through the system, only the load balancer could be public, every other service and data store had to sit in private subnets, secrets had to be managed centrally, and data and configuration had to be scoped per tenant throughout.

Solution

Cloud Combinator built the platform in the order its parts depend on each other: the network and data first, then the compute and orchestration that sit on them, then the AI and voice services, the public edge, and finally the external connections that point at running services.

3 channels

WhatsApp, Instagram and voice on one platform

80-100

Tenants targeted at production scale, 24/7 (projected)

5 services

Independently scaling containers on ECS Fargate

By the numbers:

  • 3 channels - WhatsApp, Instagram and voice on one platform
  • 80-100 - Tenants targeted at production scale, 24/7 (projected)
  • 5 services - Independently scaling containers on ECS Fargate
Changes

Acceptance is confirmed when the platform runs on AWS end to end: a message flows through the load balancer to the services, the managed data stores and serverless orchestration are in use, the AWS-native AI and voice services are integrated and benchmarked, and the external channels connect. The foundation is delivered as infrastructure as code with a build runbook, so Jafar can reproduce it across environments.

  • A cleanly separated foundation — A public load balancer, private ECS Fargate compute, and managed data behind it, so each part scales and deploys on its own without becoming a bottleneck.
  • AWS-native AI and voice, de-risked — Amazon Bedrock, Transcribe and Polly are integrated and benchmarked against the incumbent providers, with consolidation onto AWS as the direction of travel pending results.
  • Reliable orchestration under load — AWS Step Functions handle tool-calling with retries, timeouts and fallbacks, and AWS Lambda absorbs event-driven bursts.
  • Isolation and security by design — Only the load balancer is public, all services and data sit in private subnets, secrets live in Secrets Manager, and data is scoped per tenant throughout.
  • Handover — The whole platform is defined as infrastructure as code with a build runbook, alongside observability through CloudWatch, X-Ray and Langfuse, so Jafar owns a repeatable, observable foundation.

With the AWS foundation live and the AI and voice benchmark running, the platform is positioned to consolidate onto Amazon Bedrock, Transcribe and Polly as the results come in, and to scale the platform per tenant and per workload as adoption grows, with cost levers such as prompt caching and a Compute Savings Plan available once usage settles.

AWS Stack

Amazon ECS

(Fargate) for running each application service as an independently scaling container with no servers to manage.

Application Load Balancer

For a single, TLS-terminating public entry point in front of private services.

Amazon Bedrock

With Knowledge Bases, Amazon Transcribe and Amazon Polly for the AWS-native language model, speech to text and text to speech.

AWS Lambda

And AWS Step Functions for event-driven bursts and reliable, retrying tool-calling orchestration.

YOU MIGHT LIKE

Related success stories

View all case studies

Case Studies | Insights

Utilising Language Recognition, Speed, and Enhanced Security to Make Social Media a Force for Good

  • Here, we take a detailed look at how the Cloud Combinator team collaborated with another cutting-edge AI service provider that provides intelligent systems to “make social media more social” for brands and users alike.
  • Arwen AI is a UK-based startup specialising in AI solutions to manage and enhance brands’ social media interactions. Founded in 2020 by Matt McGrory, Dr. David Cole, and Joel Bailey, Arwen. AI focuses on using AI to automatically detect and remove spam, toxic comments, and other unwanted content from social media platforms.
  • The team at Arwen have three core products. ‘Moderate’ is focused on identifying and removing toxic content from social media channels. ‘Engage’ helps brands identify and engage with meaningful conversations on social media, and ‘Customize’ allows brands to apply bespoke algorithms to their channels - creating an even more effective moderation and engagement.
Read more
CONTACT US

Ready to turn AI into impact?

We'll help you spot the highest-value opportunities, reduce risk around your first AI initiative, and define a clear path to results from day one.

Why talk to us:

Outcome-driven recommendations

AWS-recognised delivery expertise

Risk-aware AI adoption

Clear next step, not a sales pitch

Start with a focused 20-minute conversation about your goals — no pressure, no commitment.

We use cookies

This website uses cookies to enhance user experience and to analyze performance and traffic on our website.

See our Privacy Policy for details.