Coda AI: Everything you need to know

Explore

Coda AI: Everything you need to know

What I've learned about using AI as a work assistant, and where it adds value.

Katy Turner

How can AI actually provide value in my work day?

Ever since generative AI made a big splash at the end of 2022, I’ve heard variations of this question from friends, coworkers, and anyone using productivity tools for their work. When ChatGPT was first released, I was among the many who eagerly signed up and had some fun—asking questions, creating haikus, and generating funny stories to send friends. But just like many others, I also wondered how to translate the benefits of this game-changing technology into my work day, and where generative AI could make a meaningful impact on my to-do list. If this is also a topic you’ve also been thinking about, this doc can help. I’ll provide some context on how the Coda team decided to explore this question, which not only impacted our AI-integration’s roadmap, but also provides a lens through which you can evaluate any AI-enabled productivity tool. Let’s take a look!

What would an ideal AI work assistant do?

In early 2023 when I joined the new Coda AI team as the dedicated AI Marketing Manager, they were deciding if Coda should pursue an AI-integration by first exploring what an ideal AI assistant would even look like. We interviewed executives, investors, IT admins, and everyday users of productivity tools to have frank conversations about the typical work day, and imagined how this new exciting technology could potentially play a role. As we had more and more conversations, a clear set of criteria emerged.

An AI work assistant should:

Understands the actual information in your company, not just what’s on the internet.

Ready available in a surface where you spend all your time, not a separate experience.

Complete tasks at scale, not just one-at-a-time.

Employees wanted an AI tool that didn’t require opening another tab, and could look at the content you’ve already created—instead of starting fresh each time. While there was a lot of enthusiasm for what AI could do, this was balanced with the desire to reduce “work about work,” and not just add another tool to the list.

We used this criteria to create a “context aware” work assistant, which means Coda AI can reference content across your doc, including tables, pages, and text. You can learn more in

Coda AI: context aware⁠

How can an AI work assistant help my team?

During our research, we also explored ideal use cases for an AI work assistant. While many shared tedious or painful tasks they complete frequently, we also heard about how getting more insights, having a virtual collaborator, or offloading an entire series of tasks could make a material difference.

After we collected hundreds of examples from teams of all sizes, we grouped these into three categories.

Writing assistance: Creating briefs or other content, editing, reviewing or summarizing text, or generating new tables.

Knowledge assistance: Answering questions, finding content, or getting specific insights from data (granular and high-level).

Task assistance: Generating a holistic AI solution that connects to other tools and can perform tasks at scale.

As we developed Coda AI and started sharing our research, alpha and beta testers told us this set of criteria helped them brainstorm how and where to apply AI to their work. For example, thinking of new places where AI can whip up a brief before a meeting (writing assistance), give insights instead of asking a colleague (knowledge assistance), and help with repetitive action items (task assistance).

As you are evaluating any AI-enabled tool, this framework can also help you create a scorecard, or differentiate between what you can use for personal productivity versus team productivity. New AI tools are emerging every day, and no two teams are alike, so having a framework makes it easier to determine how they can provide value for your work.

FAQ

How much does Coda AI cost?

Unlike other tools where AI is a paid add-on, Coda AI is included for Doc Makers. Editors also receive a free trial of Coda AI. If you’d like to learn more about pricing and usage, visit

this help article⁠

What models does Coda AI use?

Right now, Coda AI is leveraging OpenAI and we’re continuing to test multiple models to determine what provides the best user experience. We are also exploring a Coda hosted model. Please reach out to us if you’d be interested in learning more about the Coda hosted model.

How does Coda AI use my data?

To deliver the Coda AI functionality, we will be sharing your information with our third-party AI providers when you use the AI features. For example, if you ask Coda AI to summarize a set of meeting notes, those meeting notes will be sent to the AI providers. We do not allow any AI third-party providers or contractors to use your data for training their models. Coda does not use the AI inputs and results from enterprise customers’ use of Coda AI for the purpose of improving our own AI functionality. Coda does, of course, separately store doc content for the purpose of providing our services as a document platform. For general information about security at Coda, you can visit

coda.io/trust/security⁠

to learn more about how we protect your data. Our

Privacy Statement⁠

also goes into further detail.

Ready to get started?

⁠

Coda AI: context aware⁠

: Learn more about why an AI assistant should be “context-aware” and how Coda AI can reference content across your doc.

⁠

1. Writing assistance⁠

: Explore how Coda AI can help with your writing no matter where you are in the process, from blank page to final touches.

⁠

2. Knowledge assistance⁠

: Get insights at any level, whether you want an answer to a quick question, or are trying to create a dashboard for your leadership team.

⁠

3. Task assistance⁠

: Set your workflows to autopilot, while incorporating data from other tools.

⁠

Resource appendix⁠

: Browse guides, docs, and other resources to get started quickly with Coda AI.

Gallery

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.