When producing everyday visual collaterals, like an internal presentation or a social media post, many organizations rely on graphic designers or visually-capable folks who use their creative chops to come up with designs aligning with the brand language. They take care of everything, starting from colors and layouts to fonts, visual elements and the spacing between them.
But what if a company doesn’t have a dedicated person for the job? Ideally, the answer would be going with tools like Canva where you pick a stock template and customize it with different ideas and brand assets— but even that requires some level of ‘designer’s eye’. Plus, moving around elements and experimenting with them also takes a lot of time.
“Most of us can relate to hours spent trying to use stock templates,” re-arranging them to produce the desired result, notes Stefan Olafsson, the co-founder of Typeset, a generative AI design engine launched this week to fix this problem with machine learning (ML) and natural language processing.
It has the potential to produce required visual assets in a matter of minutes.
While most design tools in the market require users to cram their ideas into stock templates, Typeset goes the other way around.
The user, as the company explains, just has to select the content type they want to create and start typing the information. Typeset analyzes those words in real time and pulls together different designs for that content, complete with relevant imagery.
“In the blink of an eye, Typeset automatically designs thousands of combinations, based on exactly what you’ve written or added. Incorporating hand-picked imagery and harmonious palettes, Typeset’s generative technology creates content that simply goes together and accomplishes what you’re trying to achieve,” Olafsson explains.
As a result, using the same content, a person can have different variations of different types of visual assets (from social media posts and presentations to eBooks and newsletters) in a matter of minutes. The user does not need to worry about reformatting, resizing or moving around elements.
In an exclusive demo, the tool was able to generate an instant presentation of AI coverage at VentureBeat by pulling stories and their summaries from the author’s page. It produced multiple variations of the presentation, with images and content aligned in different ways.
As a different variation was selected, the content and the visual elements were adjusted according to it. Meanwhile, the cover image for the presentation was based on the title and pulled from Typeset’s internal library.
“The core design engine in Typeset relies on a heuristic rule engine to achieve consistent design, following the same decision path a designer would when designing or laying out visual content. We use common natural language keyword extraction to assist in choosing imagery, and offer contextual OpenAI text generation (through an AI mode) depending on the type of text being written,” Olivia Herron, GM of innovation at Typeset, told VentureBeat while giving the product demo.
She added that the images mixed in with the visual assets generated are mainly offered via rights-free image libraries, and not generative AI models.
Typeset was originally founded as “Dropdeck” a few years ago by Olafsson and his Lucidworks colleague Dr. Bjarki Holm.
The duo focused on solving the design problem but the offering largely remained in stealth. Earlier this year, it was acquired, rebranded and enhanced (with support for more content types) by the creator-focused e-commerce platform Samcart.
In early pilots conducted following the launch, Typeset saw 100% month-over-month growth in its paying subscriber base and over 10,000 pieces of content created to date. Currently, the customer base stands at over 3,900 – 40% of which are businesses.
Herron also noted that Typeset is suited for companies of all sizes, with monthly plans costing $19. In fact, the design tool is already being used by startups and Fortune 500 companies working in segments such as advertising, consulting, retail and real estate.
“Plans are still being developed, but the vision (of Samcart) is (to create) a comprehensive ecosystem for digital creators to create, build, and sell online,” Herron added. Typeset remains core to this strategy.
With the official launch, Samcart expects Typeset to draw even more traction. The company plans to expand the product with new features and content types. This includes new team collaboration features, like version control, and support for social ads and imagery, infographics and additional themes and layouts. The team-specific features, however, are likely to come with an enterprise or team-tier debuting at a later stage.
“We’re continuing to focus on ensuring that our output is consistently beautifully designed content,” Herron said.
Head over to our on-demand library to view sessions from VB Transform 2023. Register Here
When producing everyday visual collaterals, like an internal presentation or a social media post, many organizations rely on graphic designers or visually-capable folks who use their creative chops to come up with designs aligning with the brand language. They take care of everything, starting from colors and layouts to fonts, visual elements and the spacing between them.
But what if a company doesn’t have a dedicated person for the job? Ideally, the answer would be going with tools like Canva where you pick a stock template and customize it with different ideas and brand assets— but even that requires some level of ‘designer’s eye’. Plus, moving around elements and experimenting with them also takes a lot of time.
“Most of us can relate to hours spent trying to use stock templates,” re-arranging them to produce the desired result, notes Stefan Olafsson, the co-founder of Typeset, a generative AI design engine launched this week to fix this problem with machine learning (ML) and natural language processing.
It has the potential to produce required visual assets in a matter of minutes.
Event
VB Transform 2023 On-Demand
Did you miss a session from VB Transform 2023? Register to access the on-demand library for all of our featured sessions.
How exactly does Typeset help?
While most design tools in the market require users to cram their ideas into stock templates, Typeset goes the other way around.
The user, as the company explains, just has to select the content type they want to create and start typing the information. Typeset analyzes those words in real time and pulls together different designs for that content, complete with relevant imagery.
“In the blink of an eye, Typeset automatically designs thousands of combinations, based on exactly what you’ve written or added. Incorporating hand-picked imagery and harmonious palettes, Typeset’s generative technology creates content that simply goes together and accomplishes what you’re trying to achieve,” Olafsson explains.
As a result, using the same content, a person can have different variations of different types of visual assets (from social media posts and presentations to eBooks and newsletters) in a matter of minutes. The user does not need to worry about reformatting, resizing or moving around elements.
In an exclusive demo, the tool was able to generate an instant presentation of AI coverage at VentureBeat by pulling stories and their summaries from the author’s page. It produced multiple variations of the presentation, with images and content aligned in different ways.
As a different variation was selected, the content and the visual elements were adjusted according to it. Meanwhile, the cover image for the presentation was based on the title and pulled from Typeset’s internal library.
“The core design engine in Typeset relies on a heuristic rule engine to achieve consistent design, following the same decision path a designer would when designing or laying out visual content. We use common natural language keyword extraction to assist in choosing imagery, and offer contextual OpenAI text generation (through an AI mode) depending on the type of text being written,” Olivia Herron, GM of innovation at Typeset, told VentureBeat while giving the product demo.
She added that the images mixed in with the visual assets generated are mainly offered via rights-free image libraries, and not generative AI models.
Early days and subsequent growth
Typeset was originally founded as “Dropdeck” a few years ago by Olafsson and his Lucidworks colleague Dr. Bjarki Holm.
The duo focused on solving the design problem but the offering largely remained in stealth. Earlier this year, it was acquired, rebranded and enhanced (with support for more content types) by the creator-focused e-commerce platform Samcart.
In early pilots conducted following the launch, Typeset saw 100% month-over-month growth in its paying subscriber base and over 10,000 pieces of content created to date. Currently, the customer base stands at over 3,900 – 40% of which are businesses.
Affordable for enterprises of all sizes
Herron also noted that Typeset is suited for companies of all sizes, with monthly plans costing $19. In fact, the design tool is already being used by startups and Fortune 500 companies working in segments such as advertising, consulting, retail and real estate.
“Plans are still being developed, but the vision (of Samcart) is (to create) a comprehensive ecosystem for digital creators to create, build, and sell online,” Herron added. Typeset remains core to this strategy.
With the official launch, Samcart expects Typeset to draw even more traction. The company plans to expand the product with new features and content types. This includes new team collaboration features, like version control, and support for social ads and imagery, infographics and additional themes and layouts. The team-specific features, however, are likely to come with an enterprise or team-tier debuting at a later stage.
“We’re continuing to focus on ensuring that our output is consistently beautifully designed content,” Herron said.
VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.
Author: Shubham Sharma
Source: Venturebeat
Reviewed By: Editorial Team
