Question Time: securing interview data

The Problem

Interviews form an important part of any investigative process. They become valuable digital assets that require care and attention in the form of appropriate management, access control, transcription, storage and retrieval.

One agency was looking to better store and manage interview data securely via a new data platform, with machine-learning based automated transcription.

The existing system required administrative personnel to transcribe interviews manually. With a high degree of variation in audio quality and an average interview length of one hour, the process was time consuming and expensive. A backlog of transcriptions would quickly build up, as an hour interview could take five to 10 hours to transcribe, depending on the quality of the recording.

Search and retrieval times for interviews were also an issue due to the legacy storage practices, using physical media and warehouses.

Kablamo had the cloud digital asset management platform that could solve these issues. Throughout this project, the cloud platform proved it could serve as a critical component in a future which unlocks the true potential of interview data.

Our Approach

The project equally prioritised speech-to-text optimisations, transcription editing, storage, retrieval, encryption and security. The Kablamo-built DAME (Digital Asset Management Experience) was the tool to meet these priorities, as it combined organisational-centric design and AWS machine learning capabilities. DAME could be developed into a standalone version tailored specifically for the needs of the agency. The customisation ranged from finite requirements such as custom vocabulary to high-level security practices, such as log auditing and encryption.

The Kablamo team spent time with users to observe the current processes and to establish user journeys for the collection, storage and transcribing of the interviews. It became clear that a transcription automation and editing platform was required so staff could produce transcriptions as efficiently as possible.

Initial prototype of the new cloud based data platform and editor, delivered in less than two weeks weeks, was given to agency teams to trial, with user feedback aiding the optimisation of a finalised bespoke DAME implementation. As part of the process, AWS Transcribe was tested with real interview data of different qualities and varieties - including accents, slang usage, diction, recording equipment, and file type. Success rates were monitored, with the results informing the development of the machine learning solution.

Time Saving
Automated process to reduce admin time
Improved Productivity
Increased efficiency and quality of outputs
Peace of Mind
Assets secured with encrypted cloud-based storage

The Result

This agency was able to explore and demonstrate the benefits of machine-learning based speech-to-text transcription. Full-scale encryption across this database means that sensitive information is securely stowed. With the benefits clear, there is now a roadmap to a cloud-based, digital asset management system which is intuitive, secure, and meets the needs of the organisation.

The editing platform proved easy and intuitive to use, thanks to a focus on designing the platform’s user experience. In the future, the agency can explore additional tools (such as foot pedals) to add new efficiencies to the digital asset management system. The speech-to-text automation was shown to reduce administrative staff’s workload by 50% or more for most audio recordings.

Machine learning solutions were specially trained to recognise some regular phrases and local place names, remove mutters, and to apply grammar rules to common phraseology and colloquial abbreviations such as “gonna” (going to).

This solution goes well beyond transcription efficiencies. It opens up a powerful machine learning based future that can support investigative outcomes. In the near future, disparate interview data can be securely meta tagged, searched and linked as the new platform continues to evolve. It might even create the opportunities for investigators to blow cold cases wide open by making connections that would have previously remained invisible.

DAME goes far beyond delivering transcription efficiencies. It lays the groundwork for a “living” interview solution that could actually support positive investigative outcomes in the future.

Design Lead, DAME