Creating Seamless Viewer Experiences and Driving Higher Media Production Efficiency with AI

Posted on January 4, 2020

4 Minutes Read

Client Profile

A top 10 multinational mass media conglomerate that is well known for creating entertainment experiences that drive conversation and culture. With more than 170 networks that reach over 700 million global subscribers, this American media company is giving audiences across the world access to their favourite entertainment. `

The Challenge

To meet the growing demands of its viewers and drive better monetization opportunities,  the client wanted to deliver 3 years worth of TV show content to various OTT media-service providers like Netflix, Hulu, Amazon, etc.  In order to do this effectively, each third party network required every content package to be compliant with its guidelines and standard specifications. If there was even a single error and the video did not meet a specific criteria, the entire package was rejected and sent back to the production team for rework.

With its existing video workflows, this meant the production team had to spend a lot of time and effort to manually identify problem areas within each video before making the creative edits. These manual editing processes were repetitive and often took a toll on the employees – eating away valuable bandwidth which could have been utilised for more creative and subjective tasks.

Due to the longer turnaround times, fewer files were processed. This not only put a strain on the time to market, but it also drove up workflow expenditure since more full-time employees were needed to carry out redundant tasks.

With this current set up, one editor was able to process a maximum of 12 files a day.

Some challenges the team faced:

  • Repetitive tasks eating into bandwidth that should be focused on creativity
  • Manual editing and QC within content assets were tedious & error-prone
  • Higher committed costs due to lack of workforce scalability
  • None of the data generated was captured for acceleration or AI-based future-proofing
  • Failure to adhere to the compliance standards for OTT and international distribution

To overcome these challenges the client was looking for an intelligent solution that could help the production teams deliver content faster in a more compliant way and give employees more time to focus on creativity.

The Solution

AthenasOwl Smart Workflows, an AI-assisted solution helps media companies streamline and automate their media workflows.

Earlier, editing each file would take the production team at least 45 mins to process. They spent a large part of their time synchronizing the misaligned master video files to the proxy audio files. On average, identifying misalignment within videos typically took nearly 23 minutes (that is 50% of the total time spent on just QC)

With a pre-trained custom AI model performing video similarity, it helped the client automate almost 100% of problem identification and up to 65% of redundant and manual editing tasks, thereby increasing their overall operational efficiency by 5X. So, instead of editors wasting nearly 50% of their time trying to identify problem areas within each file, the AI model was able to pinpoint the errors accurately, so they could focus on fixing alignment issues and glitches within the file.

This brought down the processing time from 45 minutes to just 6 minutes per file. So, now a single creative editor within the team can process up to 90 files in a single day! (from a mere 12 files earlier)

By training AI systems to take over 80% of its repetitive manual reliant tasks, AthenasOwl Smart Workflows helped the client boost its efficiency by 5x.

Business Impact

An Auto AI platform helped augment the client’s existing workflow by dynamically training a base model to make decisions that were previously being done manually. The result was an optimized post-production workflow that:

  • Enabled more control over workflow expenditure: With a pay per use model, they can pay only for the assets that are processed without committing costs.
  • Reduced redundancy: Saved the team from repetitive efforts and time wasted identifying content and mapping it to its respective derivatives.
  • Narrowed searches to help the post-production team fix problems more efficiently: By automating the anomaly detection process to ensure compliance with IMF/OTT standards, they can now jump directly to glitches and fix them quickly. Gone are the days of watching 30 minutes of video to find 5 seconds of issues.
  • Freedom for creativity: The team was now able to focus on more strategic & subjective tasks like making creative edits to the video,  by automating repetitive, objective tasks.