Blog

Imagine a design environment where there are no defined rules to follow and infinite experimentation is encouraged. This is the space in which Tome’s product design team works as they build a new kind of 

 with AI for millions of professionals around the world.

Product designer at Tome, Shannon Rhee, shares her thoughts on designing a novel AI product for the workplace where large language models (LLMs) are taught to generate endless types of outputs, and how to navigate a design world with infinite possibilities.

Tome: From a design standpoint, what’s been most challenging, or surprising to you, when working with LLMs?

: When you’re designing for a nondeterministic system, you have to relinquish some level of control and precision that you’d otherwise have with traditional product design—how do you teach an LLM what a good layout looks like, what typographic rules exist, and what best practices to adhere to?

These are not easy questions to answer and are exciting challenges that the Tome design team are tackling on a daily basis.

Tome: As a product designer, how is AI changing the way you work?

: It’s not only changing the way we work with AI, it’s also changing how we 

 AI. In terms of visual and product design, it’s all about how we build the software and also think about the visual and experiential aspects of Tome specifically. What do the interactions between the UI and the UX look like?

Shannon Rhee

Product Designer, Tome

We think about collaboration in terms of understanding what’s possible and how it transforms Tome’s product. There’s a level of serendipity that we constantly deal with, because with AI, you’re never 100% sure of the outputs you’ll get. But that’s also liberating and where the innovation comes in.

Tome: What are some specific things that you and the team keep top of mind when it comes to weaving in principles of design with AI?

: There’s a certain element of surprise that you learn to get more comfortable with as a designer when working with AI. We’re guiding the LLM to generate dynamic layouts by teaching it the purpose of specific layout types.

Designing for AI requires acknowledging that the way a user engages with the product can yield significantly diverse outcomes — we have to embrace and design for flexibility. We analyze why things might be happening and make small adjustments and then we iterate and evaluate over and over again.

We’re playing with optionality and the levers we have to control the output—It’s like adding a bumper to lanes in a bowling alley; you put in guardrails and provide instructions to guide it.

Tome: Anything you’re most inspired by when it comes to AI and design at Tome?

: It's a fun time to look for inspiration in AI products today. Because the space is evolving so quickly, design paradigms have yet to be established in a formal way. We’re seeing new workflows for AI actions and existing patterns being re-purposed for new interactions.

The excitement of being a part of Tome’s design team stems from our persistent, design-centric engagement in machine learning experimentation. As a product design team, we’re constantly evolving our thinking and approach as the space matures.

For more on how Tome is pushing the limits with AI and machine learning, check out a conversation with our engineering leaders on the 

to get the latest in product news and features, and as always—drop us a line at 

A Tome Product Designer Navigates Building with AI's Infinite Possibilities

As AI becomes a more common professional tool, how can engineering teams make AI even more useful within the workplace? Tome’s engineering leaders had this top of mind as their teams explored new ways to give millions high-quality AI experiences, which led to 

We sat down with our head of AI, Ves Stoyanov, Archana Sankaranarayanan, engineering director, product engineering, and Sarah Sachs, engineering director, machine learning and infrastructure, to learn more about the tech behind generative presentations V2 and how it’s pushing the limits of generative AI at Tome.

Tome: What were the primary elements in our original generative presentation experience that the team wanted to update?

: There were a couple key elements that were top of mind when we set out to revamp generative presentation. First, we knew it was important to deepen interactive editing and advance the capabilities of the toolbar. 

Second, we wanted to give users more control in building their presentations by giving multiple ways to direct page information and design. People now have additional control and versatility in their narrative output.

Tome: How was machine learning used to accomplish this?

: We took a more advanced approach to working with LLMs (large language models). We needed to help train the model on the fundamentals of what powerful expression and compelling communication looks like.

We also thought about how to separate the actual presentation content creation from how you present the content. What are different presentation layouts that would be compelling? What combination of tiles and objects can we put together in the presentation structure that would enable better storytelling?

Asking for a range of strong approaches to page creation and content creation instead of relying on one deterministic information structure was vital.

Tome: Let’s zoom out a bit from generative presentations. Why is generative AI and machine learning at Tome particularly intriguing  and key today?

: What’s really exciting is that both are core to the entire infrastructure here at Tome. We are building things where every feature can be ingested by our AI systems, and be easily maintained. That is rare. There are many legacy institutions out there where to build like we do, it's just harder. Tome is uniquely positioned to leverage AI and machine learning at any point of the creation and honing process.

: We’re pushing the limits of LLM engineering and simultaneously deepening the work on our own models. We’re powering pages, and soon will be powering presentation generation with our own models. That’s an opportunity to further hone and refine narrative outputs, which can make the AI experience for people faster, more reliable and more seamless.

We also use a proprietary structure to communicate with the LLM—we call this format, TIF. That's a way for us to communicate narrative structures back and forth between the LLM and Tome’s back end.

: Generative AI and machine learning is fundamentally changing how humans interact with computers. The opportunity to continue expanding, exploring and pushing the boundaries of what we can do with engineering and in Tome, because it’s an AI-native platform, is incredibly exciting.

As both Ves and Sarah have mentioned, building the tech for even more interactive and iterative co-creation at Tome is a very unique opportunity. Generative creation in the tool is just another illustration of why it’s so powerful to create software fundamentally rooted in AI and ML, where all our systems can understand and converse with LLMs.

Tome: Thank you all for taking the time to share your thoughts with us. We’re excited to see what’s next from your teams!

For more tips on how to create your best work with generative presentations V2, check out our 

You can also get additional thoughts from 

 on all things AI and engineering here at 

As always, drop us a line with any questions at 

Training Models at Tome: How Engineering Leaders Push the Limits of Machine Learning & AI

Color is an essential expression vector in visual storytelling. Sometimes we want our stories to feel bold and cinematic (the inspo behind Tome's original “Dark” theme). Or, we might want colors to feel playful, reflective, or match a brand book.

After we launched our beta in March, many of our early users told us they wanted a wider range of color options, as well as an easy way to use their own brand colors. We didn't want color application to be the painstaking hassle they were used to.

So a few months ago, we set out to build a themes product that would let you instantly apply your brand colors across a tome or make your own theme. And we wanted your theme to automatically look 

If you'd like to dive in deep, here's the tome we used to kick off the project: 

1. Color shouldn't distract from the content. 

We don't think color should be a random wallpaper or compete with your narrative. It should always complement or spotlight your ideas.

2. Achieving color harmony shouldn't be time-consuming or hard

. It should be frictionless. And you shouldn't need to be a color theorist to build your own magical, harmonious themes.

Our now-live themes product hinges on both ideas. Now with themes, you can apply brand colors across your tome with a click. And if you want to create a new theme, our magic color assistant, the Auto toggle, lets you instantly make an attractive, legible theme. (We'll dive into how it works in a sec).

The result: you never have to worry if the colors across your tome's text, background, and UI jive. Most importantly, 

your story looks put together no matter what.

Jesse C

Product Design

Tome gives you full control over your theme and color choices. But there's no reason you should have to make every decision manually if you don't want to—that's where the Auto color toggle comes in.

When flicked on, Auto color makes sure every colored element on your tome, i.e., a page background, text, embed, a table, or your workspace—looks beautiful and easy to view. You just make one choice—your page color—and Tome takes care of the rest.

When you choose a page color, Auto color is turned on by default. Your tome's canvas, heading, and text colors automatically adjust to jive with your background.

You can save any theme you like to use again. Or if you have a brand theme, you can create it using the eye dropper or Hex values and save it, so anyone in your workspace can use it too.

If Tome's Auto color picks don't fit the bill, you can always veto them with the colors of your choice.

With just a click, you can apply any saved or preset theme to update your whole tome.

Behind Auto color is an original color algorithm made by Tome's design team. That color algorithm operates in what's called the LCH color space (more on that soon). When a page color is chosen, Tome makes a series of decisions on the fly to output a color palette that prioritizes:

Monochromatic cohesion (making sure key elements share an underlying hue)

LCH stands for “Lightness Chroma Hue.” This color space (aka a fixed range of possible colors) differs in some key ways from other color spaces like sRGB, which is historically used in design tools. For Tome, LCH was a better choice for themes because:

LCH is perceptively uniform, which means that modifying any of its three variables by a certain amount will always have the same perceptual color difference to the human eye.

Most monitors can display far more colors than what the sRGB space covers. LCH lets us tap into a wider range of colors, which tend to be vivid and beautiful.

When a page color is chosen, it's converted to LCH. Based on its inherent chroma and luminance, the base color is then used to generate complementary title, body, and canvas colors.

Depending on where a base color falls in the LCH color space, 

 are tweaked to achieve a monochromatic theme that's perceptively uniform, easy to view and read, and just looks great.

An important, but less obvious part of themes' magic is making sure that Tome's UI and native elements like tables are legible with your colors. So, when you adjust your theme selection, Tome's UI will automatically switch between light and dark modes to provide enough contrast between the UI and your tome's colors.

Tome's UI controls are based on an assessment of page color and relative luminance, a linear measure of light that's weighted for human vision. If the luminance is above 30%, Tome switches to a light UI. If it is below 30%, users will see a dark UI.

Because UI components sit on top of the theme surface, the change in UI treatment ensures there's sufficient contrast and legibility between the theme colors and the UI controls.

Originally, Tome's only theme preset was our Dark theme, which uses a single color scale optimized for darker UIs. With the launch of themes, we created a light mode UI to support the use of lighter colored themes. The light and dark UI modes are comprised of two elements:

: used mostly for UI elements (i.e., inputs, buttons, icons)

: used mostly for text and large surfaces (i.e. modals, dialogs)

Thanks for checking out custom color & themes! 🎨

We hope themes give you a more authentic, satisfying range of expression as you craft your narrative. We can't wait to see what you create.

Blog

A Tome Product Designer Navigates Building with AI's Infinite Possibilities

Training Models at Tome: How Engineering Leaders Push the Limits of Machine Learning & AI

The Magic Behind Tome's Color Algorithm and Themes

Craft your next great idea.