Modernising a National Broadcaster's archives

The Problem

With a primary search system that dated back to the 80s, and several siloed systems holding more than 11 million hours of video and audio files, the ABC (Australian Broadcasting Corporation) wanted to change the archive search experience for their content makers and internal users.

There were multiple warehouses across Australia filled with physical reels, several disconnected metadata systems, and five separate on-premise systems storing the content that had not yet been digitised. These legacy systems were costly to maintain and almost impossible to scale.

Previously, any content categorisation or archive retrieval was done manually. The process of finding content would take several weeks as it typically required a manual search of at least three databases. Kablamo envisaged a new cloud digital platform and search experience to unlock the value of the broadcaster’s archive media, and we would use machine learning to get there.

Our Approach

By choosing a custom designed and built cloud solution, the broadcaster removed the costs and constraints that come with traditional Media Asset Management systems. The bespoke archiving solution meant targeting precise workflows, efficiencies and user experiences that suited the customer needs.

With user-experience and design front-and-centre, Kablamo collaborated closely with stakeholders to deliver. The new digital content and archiving platform was built with serverless AWS cloud technology, which also gave a platform for developing Machine Learning capabilities.

The backend data platform has reusable components and powerful AWS cloud services to achieve cost-effectiveness and multipurpose scalability. Whilst managing the ever-growing database, it can also handle added machine learning features developed through Kablamo’s proof of concepts.

AWS Transcribe was successfully implemented in 2020. Now, audio and video upload can occur without manual metadata tagging. As well as automatically identifying key tags, it is helping producers find the perfect content quickly and easily.

Creating new
Replaced five legacy systems with one cloud-based data platform
Time saving
Reduced archive search time from weeks to milliseconds
Intuitive product
Powerful search and editing interface that was built with end users in mind

The Results

Kablamo delivered a successful cloud archive search concept within six weeks. The new, cloud-based digital archive solution was then delivered into production within 12 months. Kablamo deployed serverless relational databases for availability and resiliency that can manage the huge amount of daily content produced. The five legacy storage systems were made obsolete, with the content migrated to the new AWS cloud platform. This enables simple and infinite scalability while also delivering remote access and significant ongoing cost savings.

Since delivery, the broadcaster is now able to search, access and edit highly relevant content, much faster than they ever imagined. Employees now securely access the platform to self-search, to add relevant meta data and to manage their own personal collections. Archive search time is now milliseconds.

The solution deployed by Kablamo has provided the broadcaster with a scalable platform that can be enhanced to take advantage of AWS machine learning capabilities such as AWS Transcribe, with AWS Rekognition and Comprehend waiting in the wings. In the platform’s first six months, nearly two million archives had been uploaded while content had been processed or downloaded more than two billion times.

“Kablamo has been a fantastic key partner for us on the Content Digital Archive project.”

Head of Content Management, National Broadcaster
Images from top: Still from silent film, The Sentimental Bloke, 1919 (NFSA); Betty Cuthbert at the 1956 Olympics (WWoS); Gough Whitlam outside parliament, 1975 (NLA); Workers join the arches of the Sydney Harbour Bridge, 1930 (SLM)