A custom “Story as a Service” WebRTC solution with a scalable infrastructure using the Amazon Chime SDK.
Using a built-in AI Director and StoryStrip™ technology, Orson’s video storytelling process is fully automated and scalable, allowing for powerful narratives and story-driven video to be created through interviews with people.
Orson integrates with partner platforms and communities to create product extensions and new revenue streams that increase user engagement and create deeper human connection.
Real-time recording is different from regular recording. WebRTC.ventures understood the challenges and used their industry knowledge to recommend and implement the perfect solution for our needs. They were easy to work with and delivered quality product – on time and on budget.
The client was using a paid solution for video recordings. Moving to an internal solution would allow them to have more control of its functionality and, of course, manage ongoing costs.
They wanted a scalable interview tool that not only performed the basics of capturing streams, recording, and sending media, but also allowed for custom functionalities such as speech to text recording and other client-side features to be built at a later date.
With Orson’s core product already in AWS (and the fact that WebRTC.ventures is an Amazon Chime SDK Integration Partner), the Amazon Chime SDK was an easy choice for building a flexible, scalable, and easy to integrate SDK.
Pre-call connectivity checks
Role-based access control
Store recordings in the cloud
Scalable WebRTC infrastructure
We chose TypeScript/React thanks to its better tooling and IDE support, enhanced developer experience and growing ecosystem.
PostgreSQL was the right choice for the database thanks to its combination of reliability, performance, scalability, and advanced features.
Our partner, AWS, provides the cloud services needed for this project and on top of that it offers these in a reliable and resilient manner:
- Amazon Chime SDK with Media Capture Pipelines for real time capturing and recording
- AWS CloudFront for static content distribution
- GitLab CI for CI/CD
- AWS Elastic Container Service (ECS) for scalable dockerized backend.
- AWS S3 for files and media storage
- Terraform for Infrastructure as Code
- AWS CloudFormation & Bash for GitLab Runner instances
- Amazon Route53 for DNS management
- AWS Certificate Manager (ACM) for TLS certificates
- Amazon Elastic Container Registry (ECR) for scalable docker images storage used in ECS
- AWS Systems Manager Parameter Store and AWS Secrets Manager for secrets and parameters management used in ECS
- Amazon Relational Database Service (RDS) for PostgreSQL database
- Amazon Simple Queue Service (SQS), Slack Chatbot, Amazon CloudWatch and Amazon EventBridge for notifications
- Amazon CloudWatch and Amazon EventBridge for logging
- Amazon Elastic Load Balancer for ECS application
- AWS Lambda for additional functionality
Type of Application:
Automated video storytelling.
Recording/interview tool, audio and video stream capture, audio forwarding for speech to text, auto reconnection, auto scaling and pre-call connectivity checks.
How We Helped:
Orson, a generative AI platform that facilitates engagement through video storytelling, was looking to move away from a paid, inflexible video recording service. We built them a custom “Story as a Service” WebRTC solution with a scalable infrastructure using the Amazon Chime SDK.