Muoro Built a Custom Media Transcoding Engine on Azure

Our client runs a fast-growing digital platform with a large volume of user-generated content.

As part of their cloud strategy, they decided to migrate media processing services from AWS to Azure. However, Azure’s built-in transcoding features lacked the automation and robustness required for their platform.

They needed:

• A reliable, cost-effective media transcoding system on Azure

• Feature parity with AWS (e.g., screenshot generation, error handling)

• Full control over workflows to support long-term scale and customization

The client aimed to:

• Migrate away from AWS without losing critical media functionality

• Customize workflows for their unique video/audio processing needs

• Automate screenshots and support multiple output formats

• Reduce cloud service costs by building in-house orchestration

• Ensure minimal downtime during and after migration

Azure lacked critical media services that the client relied on with AWS:

Azure's media stack didn’t support automatic screenshot capture, which is critical for content previews on the platform.

The client needed custom logic to handle processing queues, retries, and file conversions — which Azure didn’t offer out-of-the-box.

Switching platforms without a fallback would impact users if workflows weren’t fully reliable and tested.

Muoro deployed a focused Ruby and Angular team to build a custom transcoding engine on Azure, feature-matched to AWS, but fully owned by the client.

We developed a set of Ruby scripts that:

• Managed transcoding tasks (video, audio, screenshots)

• Orchestrated uploads/downloads using Azure Blob

• Built a modular structure to add or change media rules quickly

We added retry logic, failure notifications, and rollback mechanisms to ensure:

• Fault tolerance during transcoding

• Predictable processing across edge cases

• Better logging and debugging visibility for the client

Using FFmpeg + Ruby, we enabled screenshot capture at key timestamps, giving users accurate content previews across devices and resolutions.

To build a reliable, custom media engine on Azure, we used:

1. Ruby on Rails – Backend system for media logic and orchestration

2. FFmpeg + Ruby Scripts – Video/audio conversion and screenshot automation

3. CRON Jobs – Scheduled batch jobs and health checks

1. Azure Blob Storage – Hosted raw uploads and processed outputs

2. Angular – Handled front-end views and media display logic

We delivered a purpose-built system that met all migration and media goals:

Rebuilt transcoding services from the ground up, with no drop in functionality.

Error handling and retry logic made processes more resilient and easier to manage.

No reliance on third-party tools, screenshot generation, format support, and processing logic were all owned by the client.

Avoided the need for expensive AWS services, all processing ran directly on Azure, tuned to the client’s needs.

The client gained:

• A media transcoding solution aligned with their Azure-first strategy

• Control over workflows and configurations for better scalability

• A robust system designed by a Muoro engineering team across backend and frontend

• A long-term cost advantage by eliminating external dependencies

Muoro builds custom engineering teams to solve complex technical gaps from cloud migration to media pipelines.

Let’s talk.