Be part of our each day and weekly newsletters for the most recent updates and unique content material on industry-leading AI protection. Be taught Extra
Serviette is an AI-based software for visible storytelling for enterprise folks to liven up their displays utilizing generative AI.
It comes from Silicon Valley AI veteran Pramod Sharma and Jerome Scholler, who began Osmo greater than a decade in the past to make use of AI to deliver youngsters’s video games to life on iPads. I coated their rise and exit, and Sharma got here again to me to do an interview on his subsequent huge concept.
Why is it known as Serviette? If in case you have an concept and write it on a serviette — the traditional begin of many Silicon Valley corporations — then it can generate the artwork and presentation to go along with it robotically.
Los Altos, California-based Serviette has created a software that empowers you to clarify your concepts visually with no effort. You write out your textual content for a enterprise presentation and, utilizing Visible AI, it can rapidly flip your textual content content material into visuals and simply regulate the fashion, colours, fonts, shapes, and design to enhance your storytelling affect.
I watched Sharma do a demo and he took textual content and the software generated move charts, pie charts and extra as an instance textual content concepts. And it wasn’t simply producing one thing excellent each time. Serviette offers you the facility to edit all the things and make it excellent, he mentioned. You may modify titles, captions, and different items of an illustration to deliver it into one thing that’s extra exactly what you need, Sharma mentioned in our interview.
At the moment in beta, Serviette works along with your present workflow in Google Docs, Slides, Canva, Slack, Phrase, PowerPoint, Electronic mail, Textual content, and extra. Not do you need to wait for somebody who is aware of what they’re doing to generate prototypes of your displays. Serviette will constantly introduce new types and designs to fulfill your storytelling wants.
In an interview with VentureBeat, Sharma mentioned that all of us have concepts, however successfully speaking them and successful folks over isn’t any simple feat. As consideration spans shrink and we hit peak data overload, how can we finest talk advanced concepts? Serviette turns your textual content into visuals with out your having to consider the laborious design complexities.
Serviette is pioneering visible AI that brings affect and enjoyable into enterprise storytelling. It’s designed for the hundreds of thousands of execs who promote concepts, writers, educators, and content material creators, Serviette helps you unlock the visible potential in your textual content content material. And it’s for these of us who could also be good writers however are awful at graphics, displays or artistry (raises hand).
“With the iPhone, Apple made it dead simple for anyone to take photos. We aim to do the same thing with graphics,” mentioned Sharma. “Visuals are powerful, and we’ve designed Napkin so anyone can easily communicate their ideas through imagery.”
It’s not nearly eliminating jobs or eliminating graphic designers. Fairly, Sharma mentioned that it’s extra about making design extra accessible, enabling you to clarify your concepts with out spending hours creating the correct visible and constructing it from scratch.
The way it works
Right here’s the way it works:
- Convey Your Textual content: Paste your present textual content content material or generate textual content from an AI immediate.
- Generate Visuals: Click on the “spark icon” on any a part of the textual content, and Serviette will rework your content material into beneficial visuals. Inside just a few seconds, you’ll have quite a lot of visuals to select from.
- Customise: With Serviette’s built-in modifying software, you may regulate font, shade, form, design, and extra to fulfill your fashion or model tips.
- Embed into Your Workflows: So simple as copying and pasting, you may merely take Serviette’s visuals into anywhere you want: Google Slides/Docs, Microsoft Phrase/PowerPoint, Substack, Medium, Slack, Notion, Canva, LinkedIn, X, Instagram and extra.
- Use it in a Range of Use Circumstances: Create high-quality visuals appropriate for displays, blogs, pitch decks, social content material, newsletters, knowledge and analysis stories, buyer communications, and extra.
“As someone without design skills, Napkin is super easy and convenient. Whether I need a quick diagram for a blog post or a chart for a presentation, Napkin generates exactly what I’m looking for within seconds,” mentioned Jason Miller, COO of Sprouts.ai, in a press release. “Before Napkin, we were constrained regarding what graphics we could access. Now, with Napkin, we have a designer at our fingertips, offering unlimited access to enhance our business storytelling with visuals while achieving significant cost and time savings.”
Immediate visuals to amplify storytelling
Serviette’s huge design catalog already contains greater than 30 visible design classes, corresponding to thoughts maps, move charts, Venn diagrams, bar charts, and extra.
Serviette’s crew of NLP, AI, and laptop imaginative and prescient builders, who hail from Amazon, Microsoft’s Analysis Lab, GitHub, Google, and Osmo, is difficult at work supporting extra highly effective customizations and distinctive classes, which they are going to proceed to launch within the coming weeks and months.
Scholler and Sharma offered their earlier firm, Osmo, whose reflective AI is utilized by hundreds of thousands of youngsters and school rooms and is built-in into merchandise with Disney and Mattel’s Scorching Wheels. Byju’s acquired it for $120 million in 2019.
Inspiration
In a weblog publish, Sharma mentioned efficient communication is essential within the office, whether or not you’re pitching an concept, explaining advanced ideas, or brainstorming new methods. How do you make it possible for your imaginative and prescient is known and appreciated by others? This problem impressed the crew to construct Serviette AI.
“We all have great ideas, but sharing them effectively can be a struggle. At Osmo, an AI gaming company Jerome and I founded almost a decade ago, we delighted millions of kids by making learning fun,” Sharma wrote. “Yet, as leaders, we often found ourselves drowning in documents, 100-slide decks, and long research papers. Occasionally, we’d come across a rare visual that truly captured the essence of an idea, sparking those “aha” moments of readability. We realized that visuals are highly effective communication instruments, but they’re not often used.”
He added, “We started wondering why so few people use visuals when they’re such powerful tools for communication. Our theory is that, from an early age, we’re trained to rely on writing. Text editors make it easy for everyone to collaborate and share their thoughts. However, creating visuals is time-consuming and requires skills many of us don’t have. As a result, most people stick to writing, even though visuals can provide a clearer, more impactful perspective of an idea. This is especially true in professional settings where effective communication is very important.”
At Osmo, he mentioned the founders spent loads of time on communication, studying proposals, writing talks, creating slides, doing analysis papers, and speaking forwards and backwards.
“It’s very daunting to consume all that information. And as we were looking into it, we realized that one thing we always loved about communication was visual.”
“We realized it’s incredibly hard to make good pictures of things that you’re talking about,” he mentioned. “Most people have been using generative AI for and for social media. That seems like where everybody gravitated at first.”
However the founders thought small enterprise and the enterprise had been the correct strategy to go along with AI visuals. The graphics can get fairly advanced with interactivity. You may specify boundaries of what you need earlier than you submit the thought to the AI.
“We realized is most people know how to write but they don’t know how to create a visual,” Sharma mentioned.
However making this look simple is the onerous half.
The crew is small with 11 folks they usually spent greater than a yr and a half engaged on instruments.
“We realized that most people don’t have the layer of creativity or the time to come up with something,” Sharma mentioned. “Very few people can actually start from scratch. And only in our third generation did we realize that what people really want is AI.”
“It’s extremely hard. We are building something almost like a visual engine. I give you bunch of information, and you create the right graphic and dynamically create graphics,” Sharma mentioned.
Fairly than produce wild and imaginative artwork, Serviette focuses on visuals for data. It’s not Salvador Dali.
“We’re trying to complement the written content,” Sharma mentioned.
The primary model of an AI software got here within the type of a sketching software to make easy diagrams along with your mouse. Nonetheless, the crew rapidly found that even with the perfect AI serving to within the background, customers don’t need to sketch reside with a mouse.
Within the second model, the creators launched a database of visible frameworks like Venn diagrams and pie charts. Customers might give attention to discovering a visible to make use of that matched their concepts, as an alternative of sketching. However they found that many individuals don’t know which visuals to make use of or use them successfully, even when they’ve entry to them.
Lastly, the introduction of Massive Language Fashions (LLMs) supplied the right complement to the crew’s efforts. They targeted on remodeling textual content into visuals robotically, assembly customers the place they’re most snug – writing textual content. This marked the delivery of the Serviette AI you see at the moment.
Wanting Forward
Simply because the iPhone made it simple for anybody to take high-quality photos, Serviette AI’s mission is to make visible communication simple and accessible to everybody, Sharma wrote.
“We understand it’s a really hard problem, and we have just started, but we are excited about the progress we’ve made so far. We believe we can achieve our mission by focusing on two core principles: deeply understanding text content and generating a wide range of relevant and high-quality visuals,” he wrote.
Accel and CRV are the buyers within the challenge. You may write them at crew@serviette.ai and be part of the beta at serviette.ai.