Categories
Founder

Revolutionizing Productivity: The Genesis of Gistify

Development of New AI Technologies to Identify Key
Concepts Extracted from Rich Media Content

Mark Cromack, Founder

Gistify is a revolutionary productivity tool that saves you time. Utilizing patent-pending AI technologies, Gistify takes rich media content, like podcasts, Zoom recordings, and lecture videos, and identifies key elements of importance to you. This powerful visualization acts as a “rolling abstract” of relevant words, allowing you to find what you want quickly and easily, obviating the need to always listen to or view the underlying media.

With Gistify, you get the Gist and skip the rest!

Lost Intellectual Capital – Can You Remember?

Your meeting has already been going for an hour, as the team brainstorms on key business opportunities. While ideas are flowing, the following exchange happens:

“Wait, what you just said was perfect! What was that again?”
“Ah, I’m not sure, but it did feel on point…”

Have you ever had that conversation where you heard something that really resonated yet you can’t remember exactly what was just said? Maybe in a lecture where you heard something that struck a chord, but such a fleeting point was difficult to hold on to? Or, at the end of the day, you reflect back on your meetings and calls only to wonder how many things you missed.

These types of challenges drove Cogi’s founding team to ask how we can implement tools to help ensure that a “Cogent Idea” or “Cogi” is never lost. Cogi’s fundamental mission is to create advanced technical solutions to amplify human cognitive abilities.

At first, the team was interested in capturing everything, literally. We wanted to record everything in every context to be sure that nothing was lost or missed. But then we thought:

“What’s the utility of such an immense collection of recordings without the tools to identify relevant moments? And how is relevance determined?”

So, we considered the notion of being able to inform Cogi when something was important to us. We even compared our idea to the Star Trek communicator, a wearable badge you tap when something significant is said.

Cogi’s Mobile App – Capturing Cogent Ideas

With mobile apps all the rage, it became clear that a carefully and uniquely designed mobile app could create a powerful experience for anyone trying to improve their ability to remember those relevant and important moments in their daily lives.

The Cogi mobile app provides such an experience with an exceptional interface that is always listening and standing by for those moments when you realize that something key is being said. When you hear something you want to remember, all you have to do is tap the Star Trek badge, I mean the app’s “Cogi button”, and Cogi handles the rest.

But wait… some of what was heard and relevant to what’s interesting to you is now in the past, before you even had the chance to tap a button, and that earlier content represents the trigger or cue that made you realize this concept is potentially meaningful.

Don’t worry, the Cogi app accommodates this complication by “backing up” in time to capture the context of what you felt was important. When the conversation eventually turns mundane, a second tap of the Cogi button ends the highlight, memorializing that Cogent Idea. These Cogi highlights, captured in real-time from your meetings, lectures, and phone calls, are now a collection of what you feel is important.

Visual Noise Reduction to Effectively Skim Media

We realized early on that a written transcript increased the value of Cogi highlights, as your content is now readable and searchable. And with the highlighting capability, the amount of textual content and the associated cost of creating the transcripts is much less than if you transcribe an entire recording. Yet even so, reviewing or even searching through that audio and textual content can be exhausting. Further, if crowd-based transcription is utilized, the cost associated with transcribing even just highlights can be significant.

So, our new problem statement was multifaceted.

We know that the construction of a transcript will provide demonstrable benefits, and perhaps with search, this tool can provide significant utility. And if that solution uses automatic speech recognition (ASR) technology, we can obviate much of the cost associated with these transcripts.

“But if speech recognition is fraught with word and punctuation inaccuracies, how does that provide that ideal solution we are looking for?”

Regarding speech recognition technology, even today’s best solutions vary in accuracy, both in terms of the words produced and the placement and accuracy of the corresponding punctuation. This accuracy varies based on numerous factors including background noise, loudness, speaker’s accent, and crosstalk, as multiple people speak simultaneously. Yet, the cost of these technologies is demonstrably less than human-based or crowd-based transcription. Thus, given the lower cost and faster turnaround time from automatically generated transcripts, how can we best utilize such an amazing yet imperfect technology?

Further complicating matters is that conversations and lectures are very dense. You’ll typically see anywhere between 12,000 and 15,000 words spoken for every hour of audio or video content. So that 90-minute meeting results in something like 20,000 words to search from, sort through, or utilize in some form or fashion. Much of that written content is “visual noise”, contextual information that is mostly superfluous.

The result for the Cogi team was the recognition that only a small portion of this huge array of words is of significant value to anyone participating in that meeting or lecture. Further, when you consider content that you skim, for example, skimming your textbook chapter on photosynthesis, you’re naturally keying off specific words and small collections of words to form the anchor points for meaning and relevance. Once you’ve noticed one or more of these relevant “collections” of words, you begin to drill down in the content for what you were looking for or to derive meaning or understanding.

Gistify – Get the ‘Gist’ of Your Content

Gistify is a revolutionary productivity tool that allows anyone to easily and effectively “get the Gist” (or the essence) of their important content. The Gistify service allows you to transform your media into a new, visual experience, creating an interactive tool, a “rolling abstract” of keywords and other visual elements, that allows those viewing the content to easily identify relevance and extract meaning without having to listen to or view the associated rich media content. This rolling abstract is composed of a series of abstracts, each associated with a specific section of time within your content.

Today, many of us find ourselves in online collaboration sessions. While this “new normal” is offering new efficiencies for businesses, how can we easily capture what’s important? Platforms like Zoom make it easy to record these online sessions, but few people have the time to review a 90-minute meeting. So we asked ourselves, how can we make it easy to extract the key ideas from these recordings?

With Gistify, you can revisit your meeting from last week by taking a few moments to cycle through the rolling abstract and remind yourself what the key takeaways were from that meeting. Or, if needed, you can find that action item you were assigned by your boss in order to make sure you fully understood the task at hand.



The focus on relevance identification became the genesis for the “time cloud”, a time-varying visualization derived from the raw audio or video content. The “time cloud” is a synchronized, visual map of the important and extracted elements from that media that creates a visual Gist… think of this as “Google Maps for rich media content”. Each visual element is a “landmark” (in keeping with the Google Maps metaphor) for meaningful content derived from the audio or video source. These elements include the words extracted via speech recognition, but they may also include other useful visuals like an emoji that represents the sentiment associated with what’s being said.

The Gistify service allows anyone to easily and effectively get the Gist of their important content. This includes meeting recordings from collaboration platforms like Zoom, Teams, and Webex, university lectures, phone calls, to name but a few media sources. These time-ordered visualizations are segmented to depict what’s most important for individual portions of that meeting. For lectures, you can easily sequence through the rolling abstract to find that moment from your economics class that was particularly confusing. Once there, you can either review the Gistified visualization and its corresponding contextual hints (just mouse over, or long tap on mobile, any keyword to see what I mean), or tap on a keyword to listen to or watch the original content.

Test Drive Gistify Today

The Gistify service is a transformative productivity tool, allowing anyone to easily find and extract value from their recorded content.

With Gistify, you get the Gist and skip the rest!

You can visit Gistify.ai to test drive an extensive collection of Gistified media examples across various industries and applications.

Leave a Reply

Your email address will not be published. Required fields are marked *