Ian Green - Audiovisual transcription and redaction

Intuitive design for Audiovisual data review.

Role: IC Lead Product Designer

Company: Logikcull

Cross-Functional Team: Product, Engineering, Executive team, Sales, Customer Success.

User Goal: Legal professionals urgently need a more efficient and cost-effective solution for searching, reviewing, and redacting audiovisual data.

Business Goal: To establish a leading position in the legal tech market by delivering a powerful audiovisual review and redaction feature, driving new customer acquisition and significantly enhancing existing customer retention.

Result: Logikcull now offers integrated audiovisual review with searchable auto transcription, VoiceTouch and audio redactions. One success story was a Fortune 500 company that saved $178,000 by rapidly reducing voicemail review from 600 hours to minutes.

Legal teams need a cost-effective way to analyse
the growing volume of audiovisual evidence.

Data Volume & Search Challenges:

Exponential growth of audiovisual data is a huge problem for legal professionals, often having to manually review 100s of hours of recordings.
Existing audiovisual software and review tools are complicated and require extensive training.  
Limited search capabilities within audiovisual files results in missed evidence.

Manual Process & Inefficiencies:

Inefficient manual review for critical evidence can cause delays, financially penalties or even the loss of a case. 
Lack of integration with existing document review workflows means siloed data.
The inability to quickly and accurately find information is a critical failing in today's legal environment.

Increasing costs & required resources:

Time-consuming and costly manual transcription. 
Audiovisual data review can often require large review teams which frequently leads to outsourcing to third parties.
There’s an urgent need for a cost-effective solution for searching, reviewing and redacting audiovisual data.

Driving audiovisual feature development through extensive user and market research

Start with the why? Identifying pain points and user flows.

As Design Lead, I led the research effort for this feature, encompassing user needs and broader market analysis. Through user interviews, I gained a deep understanding of users current audio-video workflows and identified key pain points. Users expressed a strong desire for the same simplicity and efficiency they experienced with our document search and review features.

User Feedback: The foundation of our design and product strategy

Our product and design decisions are driven by continuous user feedback, collected through direct customer interactions, communication platforms (e.g. Intercom), and internal team collaboration. This feedback is systematically logged and analysed in Productboard and Jira, and prioritized using the RICE framework (Reach, Impact, Confidence, Effort) to ensure alignment with business strategy.

User research

I conducted in-depth interviews and usability testing to understand user requirements, identify pain points, and gather feedback on existing solutions. I carried these out virtually via a public calendar link and in person during company off-sites.

Competitive Research

Analysed the competitive landscape within the legal market, identifying key differentiators and areas for improvement.

Market Research

Analysed existing products offering similar functionalities, identifying strengths, weaknesses, and opportunities for differentiation.

Transcription Product Research

Evaluated a range of transcription services, considering factors such as cost, accuracy, and integration with our existing systems. Collaborated with engineers to assess technical feasibility and integration challenges.

A/V Software Editing Research

Explored existing design patterns for interacting with audio and video files, such as time stamps and drag handles, to inform the user interface design.

Video Redaction Research

Investigated best practices for redacting both visual and audio content, ensuring comprehensive and effective redaction across all media formats.

User stories and requirements

As a reviewer . . .

I want to be able to quickly and easily review audio and video files.
I want to be able to carry out keyword searches across all my files (including audio and video).
I want to be able to view transcripts for audio and video files to save time.
I need to be able to produce audiovisual files and their associated transcripts with redactions.
I need the ability to select a time frame for redaction and adjust it for accuracy. 
I want to be able to filter files by type (e.g., Audio, Video).

Wireframing facilitated core architectural decisions and the development of key features like VoiceTouch.

Component Libraries vs. Wireframes: Finding the Right Balance.

In today's design workflows, utilising a component library can streamline the design process, often making it as efficient to create mockups as traditional wireframing. However, at times, I've found wireframing invaluable for its ability to isolate and focus on the core architecture and structure of the user interface, free from the constraints of specific design styles.

Aligning Teams Through Wireframes: Communicating Design Early.

During the initial stages of this project, as I began to analyse user feedback and collaborate with engineering and product teams, we recognised the need for a method to visualise potential solutions without prematurely committing to a specific design direction. Wireframing proved crucial in this regard. It allowed us to effectively communicate and agree upon key aspects, such as the optimal placement of the automatic transcript within the existing interface and the accessibility of essential interface elements like redaction buttons.

Wireframing Breakthrough: Developing the VoiceTouch Concept

One particularly impactful outcome of the wireframing process was the concept of "VoiceTouch." This innovative feature enables users to seamlessly navigate the audio/video by simply clicking on a specific word within the generated transcript. This functionality is particularly powerful when a user searches for a keyword, as the transcript highlights the relevant terms, allowing for instant playback.

Automated transcription in Logikcull unlocks searchable A/V content, simplifying review and redaction.

Creating a solution and delivering immediate value to users.

As a design leader, I'm proud of how our team streamlined multimedia evidence review. Recognizing the time burden of audio and video for legal professionals, Logikcull now automatically analyses and transcribes these files upon import, immediately simplifying the process and adding real user value.

Transcripts are seamlessly integrated within the document viewer, alongside the media. Files are instantly QC-tagged "Transcribed" for clarity.

We prioritized discoverability by automatically categorizing files and using search facets, enabling efficient filtering of audio and visual evidence. Visually, selected media files appear as scannable cards with file type, a transcription snippet, and the "Transcribed" QC tag for quick identification.

Saving legal professionals hundreds of hours in review time.

Users can search across all file types using keyword search. Now, thanks to auto-transcription, we can return search results from within audio and video files, saving legal professionals hundreds of hours in review time. 

The screens below showcase the view of an audio file in the document viewer. Users have access to the transcript and the ability to redact both the audio and the corresponding text in the transcript. This two-way redaction functionality has proven to be a game-changer for our users.

Watch the video below to see audio transcription, 'Vouch touch', and audio redactions in action.

Video UI prioritised dynamic resizing and redaction, with core features delivering immediate value.

Adapting Scope: A Phased Approach to Video Transcription and Redaction

Designing the user interface for video files presented unique challenges compared to audio. One of the primary considerations was optimizing screen space for both the video player and the transcription. Recognizing that users often require a larger video display for effective review, I explored design solutions that allowed for dynamic resizing, such as an expandable and collapsible transcript section.  

Similar to audio files, users needed the capability to redact both the audio and visual components of video files. Our initial approach for phase one involved allowing users to simply insert a blank screen with a redaction label at the desired timestamp. However, as we collaborated further with users and the engineering team, we encountered unforeseen complexities that would significantly increase the project's scope.

Although we are confident in our current design solution to achieve the desired outcome, the timeline for implementing video redactions has been extended and thus is something we continue to work to implement.The value we have delivered to users in the interim, with features like automatic transcription, search, review, and voice touch, represents a significant improvement over previous capabilities.

A/V Transcription and redaction saved a Fortune 500 company $178,000 by rapidly reducing voicemail review from 600 hours to minutes.

A Tulsa-based Fortune 500 energy company faced a daunting eDiscovery challenge: reviewing nearly 15,000 voicemail records. This task was projected to consume 600 hours of review time and incur six-figure attorney fees. However, by leveraging Logikcull's A/V Transcription with VoiceTouch, they dramatically reduced the scope of their review. Within minutes, they narrowed down their focus to just eight potentially relevant recordings, resulting in over $178,000 in cost savings.

View customer case study

Logikcull now offers integrated audiovisual review with searchable auto transcription, VoiceTouch and audio redactions.

Before:

Legal professionals often have to manually review 100s of hours of recordings. 
Audiovisual software and review tools are complicated and require extensive training.
Limited search capabilities within audiovisual files results in missed evidence. 
Inefficient manual review for critical evidence can cause delays, financially penalties or even the loss of a case. 
Lack of integration with existing document review workflows means siloed data.
Time-consuming and costly manual transcription. 
Audiovisual data review can often require large review teams which frequently leads to outsourcing to third parties.

After:

Get automatic transcriptions of audiovisual data and review them like any other document. 
Use relevant keywords to search all audiovisual files and redact the transcript and the audio file with a click. 
Filter audiovisual files using automatically applied categories in the search facet filters, for example, Audio and video.  
Stop paying for expensive third-party A/V transcription services! Logikcull’s transcriptions are automatic, quick, and accurate. 
 
Click any word or phrase in a transcript and VoiceTouch will let the user jump to it anywhere in the recording to understand the context in which it was used.
Keep their entire document review in one single place.
Utilise automatic QC tags and manual tags to find and revisit audiovisual files of interest easily.

Video redaction and advanced automated detection capabilities are under development.

Phase one of this project prioritized delivering immediate value to users by focusing on core functionalities: transcribing audio and video files and implementing robust search and filtering capabilities. This initial release significantly enhanced user experience by streamlining workflows.

Recognising the importance of delivering value quickly, we prioritized the development of audio redaction and automated PII detection in the first phase. Video redaction, while a key focus, is currently under active development and is yet to reach production.

Beyond basic redaction, we are exploring advanced capabilities, such as the automated detection and redaction of specific elements within video files, including faces and car number plates. This initiative is currently in the design and development phase.