Episode 3: May Habib on spending more time in flow

For our third episode of Working Smarter we’re talking to May Habib, the co-founder and CEO of Writer, a generative AI platform for businesses. Writer helps streamline the writing parts of people’s workflows, so they can get to the “thinking part” of their jobs more quickly.

For Habib, AI isn’t just about automating mundane or repetitive tasks, but a way to amplify the talent and intelligence of a team. There are some types of writing, such as an RFP or a JIRA ticket, where it doesn’t always make sense to start from scratch. By using AI to learn how a company works and the knowledge a person needs to do their job, Writer aims to make those kinds of writing-related tasks faster, easier, and more efficient. 

Hear Habib talk about how AI-powered tools like Writer can help increase the amount of time you spend in flow, supercharge your existing workflows, and create more space for insight, creativity, and inspiration in your job. 

Full episode transcript

Whenever I’m at work, and I’m having trouble writing, I think of an old piece of literary advice: just fill the page. 

Maybe it’s the worst writing you’ve ever done. Maybe it’s just the word “no” repeated 500 times. It doesn’t matter if the writing is wrong or bad or not quite right—your only goal is to put something, anything, down on the page. 

Trust me. It’s easier to improve a middling first draft than chase perfection from the start.

But with generative AI, who says it has to be you that fills the page? Or that it’s even a page at all? Maybe it’s a job description, a JIRA ticket, a letter to the board; an email to your team, an RFP, a year-end report; or a strategy doc, a slide deck for the all-hands, or customer support. In fact, there are so many different types of jobs—and so many different types of writing—where it might actually be more helpful to let AI take that first pass.

And not just helpful for you, but for your customers, your company, and your team.

I’m your host Matthew Braga, and on today’s episode of Working Smarter, I’ll be talking to May Habib. She’s the co-founder and CEO of Writer, a full-stack, generative AI platform for enterprise companies. Using its own in-house language models, Writer can help you build personalized, AI-powered apps and workflows that actually understand you and your business—from how you work to how you write. So you can spend more time in the flow of things, and less time staring at the proverbial blank page. 

That’s coming up next on this episode of Working Smarter.
 

~ ~ ~
 

I would like to start by asking maybe a bit of a big question or an existential question, but what do you like most about writing? 

I love the way it forces me to think through a problem. And there is just this, like, utter crispness to how your ideas get organized, right? Writing is thinking. There's just an ownership over the ideas when they have to be new and innovative that you have to take yourself—and writing is just a forcing function to make that happen.
 

What I love about writing—it's similar, it's the problem solving, but it's also working ideas through and trying to build something up from nothing. But the reason why I was thinking about it is because, as a writer, I think I'm always looking for things that will let me do more of the thing that I love most about writing. Whether that's tools, processes, things that can take away the things that maybe I don't enjoy as much. Or help me focus more on the things that I love most. 

And so, I'm thinking of all this because you are the co-founder and the CEO of a company called Writer. You make a product by the same name. How do you describe Writer to people? Like, how do you describe what it is and why someone would want to use it?

The company is called Writer because everything we do is mediated with words—all insight, all communication. Writing as we think about it—you know, putting thoughts to paper—that's really one element, and one part of what Writer does. 

We have customers who are uploading spreadsheets of market indice performance results and automatically writing market narratives, or testing that performance against a thesis that was written before. Everything we do is mediated by the written word, and so much of how you can break down processes at work, how you deconstruct what it means to have intelligence at work, comes down to reading and writing. 
 

I love that because I know that sometimes you talk to people and they describe what they're doing and they don't think of what they're doing as writing—because, you know, they're maybe not sitting down at a Word document or a notebook—but writing is still instrumental to what they have to do as part of their job. I'm kind of curious if you can take me back a little bit. When you set out to start Writer, what did you feel like was missing in the industry? What holes were there or gaps that you felt that a company and a product like Writer could fill? 

The big idea was writing is the last unstructured business process. There's no system of record for writing—but as you said, everyone across the business is writing. Writing is thinking. And we were defining what writers do too narrowly. Everybody writes, and if you could make that more efficient—if you could make it more streamlined—that's super strategic for a company.

So those were some of the initial ideas that we were playing around with in 2020, and the initial product hook as an AI writing assistant that gradually was doing more and more for a user. And, thanks to ChatGPT, interest in this technology has just absolutely exploded.

And we, for almost four years now, have been building an enterprise-grade platform to be able to deliver it to thousands of users at scale within a single company. So everything from how you manage multiple teams, multiple profiles, multiple brands, multiple knowledge graphs, use cases that sit on top of that—it's just really scalable, which is why we have thousands of users at some of our largest customers and growing.
 

I'm wondering, what are the bottlenecks? What are the parts of people's workflows that you're trying to fix or you're trying to reinvent? 

Mmm. That's a great way of putting it. I love putting it that way. 

That part of the workflow where you're digging stuff up, putting ideas together, and testing against other ideas that you have. Getting to the thinking part is what we accelerate. And so much of a workflow—whether it's responding to an RFP, or putting together a proposal, or writing up notes for a meeting—are a set of processes, of research, synthesis, communication, deconstruction, repeat.

And what's awesome about AI is for someone like me, for example, who can be distracted so easily, the ability to stay in flow by linking more closely together the thinking parts of a workflow is really cool. And we get that feedback from customers all the time that folks’ job satisfaction goes up not because we're automating the mundane, or removing repetitive stuff off their desk, etcetera, but we're increasing the amount of time they spend in flow. 

And when you talk about getting into that flow state—trying to kind of find ways to change the way that we work, or work better, work smarter—what are some of the things that Writer can do to help with that? I'm wondering if maybe you can break it into buckets—you know, Writer can do X, Writer can do Y, Writer can do Z, 

Yeah. Well, the way that we do it from a product perspective is break down use cases by vertical, by function, and then by category of AI capability. So it's amazing kind of breaking it down in this way into sort of like discreet use cases and workflows. Because as much as that's required to lay the map for people of "What can generative AI do? What does Writer do? How do you disambiguate?”—this is new space and a highly competitive space. 

I spend like 70, 80% of my day in that headspace. And increasingly, because so many customers are really leaning into the technology, we also get to talk about the fact that there is underconsumption of AI when you think about it in purely use-case based ways. Putting tools that can read between the lines and help you see things you didn't see before open up all sorts of new ways of working as well. 

But I'm going to come back to answer your question. So if you think about verticals, we're very vertical specific. So health care, financial services—and even within financial services, we're really clear. We don't touch capital markets. It's investment banking, wealth management, insurance. Even within insurance, we have sub-buckets of things that that we like to do. Retail and telecom.

And so, by function—let's say corporate functions, for example. If you break out HR, IT, finance, things like performance reviews and job descriptions and user stories and JIRA tickets and writing POs and doing RFPs—there's just so much that is pretty consistent across a vertical, but depending on what the company is, needs to be really specific. 

I'll give you an example. If you are Marriott, you are parsing through a lot of RFP responses. If you are a B2B SaaS company, you are submitting a lot of RFP responses. And so when you load up an enterprise version of Writer, needing to start from scratch on a RFP application isn't that useful, right? Our value prop is: for the verticals that we focus on, for the functions that we serve, you get a platform that is ready with the solution map for each function. And the kind of low-code, no-code approach to building these use cases is how we've differentiated ourselves. It's why folks find such fast time-to-value. 

If you go back to the matrix I was explaining—vertical, cross-function, cross-capability, with capabilities being generating, analyzing, governing—we present really complete solution maps for people where they're already 80% of the way there. They plug in their data—we fine tune on the way they write RFPs, to go back to that example, the way they review RFPs—and the focus is really on the last mile of data work, which is where all the alpha is, and the last mile of user adoption, which is how you actually get transformation. If people don't use it, or only 20% of people use it, you're not really getting the kind of impact folks are looking for.

So that's maybe deeper than you wanted to go on use cases. But, you know, this stuff is not easy and folks think they can kind of spin up a raw large language model and, zero-shot, have something that works, right? And yeah, there are some use cases where that's fine. If you're building features and folks are still experimenting, that's okay. But for deep internal workflows, like the heart of your business, there's still so much work you’ve got to do around the large language model. And we make it easy by providing this full stack solution.
 

Well, and there was a word that you used a second ago when you were using that RFP example. If I'm a company, I want to write RFPs, I can create a template. And this template that you've helped me build is also pulling from all the other RFPs that, you know, our company has written. How important is personalization when it comes to building an effective AI tool like Writer? Like, what kinds of knowledge would help Writer most effectively solve my challenges or my business' challenges? 

This is what makes us, I think, really compelling. You know, we're not taking like a boil the ocean approach. We're not saying everything goes into one knowledge graph and this is going to serve all use cases. It doesn't really work that way. So by personalization, I would cut that in all the ways that are relevant. Like, the customization of every generative AI application built on Writer to the data that is relevant for that application is what we'd want to be doing.

So if you’re using Writer for release notes and product marketing and email and customer support and FAQs, those use cases are informed by both previous examples of that work but then also knowledge that a model should have about your company to be able to form that use case. And so, the knowledge that goes in for retrieval augmented generation could be shared across use cases, right? But you are taking this fundamentally use-case based approach to figuring out what data you need to customize. 
 

And can you tell me a little bit more about the Palmyra LLMs that make a lot of this possible? 

People ask us all the time, like, “Why? Why do you build your own models? Why aren't you just plugged in GPT 4?” A lot of the accuracy that we're able to achieve is possible only because we have a lot of control over the underlying model. We're committed to state of the art models. So we're very focused. We don't do code. We're not generating images. It's what it says on the tin—it's writing. And because of that, we can really maintain that state of the art. 

So much of how you implement retrieval augmented generation—so much of how you implement our knowledge-graph capability—is dependent on that underlying model. If you've got a model for a particular use case—and we have a customer in financial services where they fine tuned a model on a ton of sentiment work across, you know, literally like 60 million users. And so there's a subset of their use cases where they've plugged in their own model. We can do that. This is like a highly composable stack. 
 

There are obviously a lot of different industries and jobs that involve writing in different capacities. And I know before you mentioned a couple of the verticals at which Writer has had the most success. You're talking finance, you're talking insurance. I'm wondering if you could elaborate a little bit on, you know, are there specific types of tasks within those industries that you've found that Writer has really excelled at? Things that have really helped transform how people within some of those verticals do the work they do?

There are so many use cases within our healthcare business from United Healthcare to CVS. There are a lot of folks who are, for example, trying to understand the research coming out, guidelines coming out, rules coming out from an NIH to the university and academic institutions that they track on hundreds of different topics.

And we see a lot of folks—it's both digesting content and research that's come out. It is producing synopses, producing abstracts, for things that they're writing. Translating it into language that lay people can understand, right? Turning it into more mainstream level insight and bulletins. So we see a ton of that in healthcare specifically, taking stuff that is just not that digestible for the common person. If you are taking a set of guidelines from the federal government around new diabetes care instructions, or new ways that some Medicare program is going to work, you now have to turn that into material that educates and maybe even changes behavior. And that's pretty complex stuff.
 

How do you get people to successfully incorporate a tool like this into their day to day? 

So that is a multi-million dollar question.
 

[Laughs]

Fifteen to 20% of people will become just natural fast adopters. And no matter how hard you try, 10, 15% just don't want to touch it. What you do with that middle part is really where the work goes. And honestly, people ask all the time, like, “How do we make this go faster? What's your advice? How do we staff this up?” It's almost always something along the lines of: take the folks who are already leaning in, who are already so excited, who are the most positive, and we're going to showcase what they build with Writer to the rest of the org. And that just helps so much, because people start trying things and it's really only when you've tried it on your own workload that you see just how much leverage you can get. 

Writer is focused a lot on the shared team workflows, right? We're not a co-pilot, like, “Go summarize your emails, Matt.” Whenever folks give me that as a use case I laugh. I'm like, who is sending you thousand-word emails? Like why is this the top co-pilot use case? I don't understand. So we have this natural advantage where, if you're at L'Oreal, the way that Kiehl's now merchandises something—that has like permanently changed. We are that workflow and that's the behavior change. That's the hook. 

But there are a lot of other use cases where there's a number of ways to do something. Showing folks the faster, easier way, and really just getting that muscle memory—nothing beats a face-to-face, 30-minute, “Matt, let's use this together. You drive. You type. You share screen.” And that's what we do. 
 

I think I've also heard you say before that part of how you bring people along is by integrating with the tools and workflows that already exist—the things that people are already using. And so I was wondering if you could elaborate a little bit on that. It's one thing to come to a company and say, "We have this brand new tool. We want everyone to download it, install it, open it up on their laptops, use it every day." But it's another thing to say, "You can keep using the things that you already use, but we're going to slip this thing into your workflow along the way.” What is the importance of that?

It's important to be able to give folks lots of options. So we have a desktop app, we have a Chrome extension, we have a Figma app, we have a web app. Every use case has an API; we can build that into workflows. So I do think generative AI has to be omnipresent to be most responsive to how folks want to do something. You can't really force the issue. 

We have a customer who calls it “workbench.” There are some workflows that don't have a workbench. Like there isn't a place where it happens. It sort of happens in notes or docs, or there isn't like a standard process. When we introduce AI to reinvent a process AI-first, like there is now a process, and that tends to be driven by API—like, stuff just showing up in a database or showing up in a CMS or showing up in a CRM that wouldn't have been there previously.
 

How do you quantify success for a tool like Writer? Like, what does a successful deployment of Writer, or like a successful Writer user look like to you? Is it time saved? Is it money saved? Is it being able to report that I feel more creative or productive? How would you define that? 

There is certainly just the quantifiable "Matt's used us a few times this week. Awesome.” There is the really fun qualitative side, like, "Oh man, Writer saved me. Oh fuck, like, I got this thing and I would have had to stay all day, or all night, or all weekend, and I got it done because of Writer.” Or, lately we've a number of people tell us that they've been able to stay in jobs longer as a result of Writer—just from a pure ADD perspective, you know? Like, “I actually get through the stuff I hate to get through to the stuff I love, like, way faster because of Writer.”

So all of that stuff is—that's what we live for, folks really feeling like we have helped them enjoy their jobs more. From an executive perspective, of course, they want to see business value. Again, we're, the smaller brand, right? We compete with these tiny companies called Microsoft and Google and Open AI to an extent. So we're really focused on kind of proving the business value. I don't think anybody asks a CIO to prove the business value of a co-pilot. They just give everybody licenses and hope for the best. That's just the market that we're in right now.
 

You used this example a second ago that I liked, where Writer helped someone with ADD do their job more effectively. And I'm wondering whether you think Writer can help level the playing field in other ways too—say, for people who didn’t learn English as their first language, or don’t think of themselves as particularly strong writers?

My co-founder is Exhibit A. He is self-taught English and, love him to death, now he writes in perfect sentences. [Laughs.] And we hear that all the time—that it is just such a democratizing effect on folks who may not work in the U.S. or don't have English as a first language. 

I think the more subtle improvements that are going to have real dramatic impact on the workplace and who we hire, who we retain, who moves up, is the fact that if you've got the right mindset, somebody who's been at a company six weeks or six months can be as good as somebody who's been at the company for six years. If we're able to successfully do what we want to do as a company, our vision is to transform work. It's to really take work from this collection of tasks and workflows, and by pairing it with knowledge, really level up the amount of intelligence inside of a company. Everything an organization knows, everybody in the business knows. 
 

When you talk about writer and that goal of being able to transform the way people work, I'm wondering what that looks like for you day-to-day. Like, how does writer factor into to your day-to-day workflows right now? 

Yeah, that's such a good question. So, I get a May's Daily Prep sheet. It's got a lot of links to it. I'm doing 15 to 25 meetings a day. It's a long day, and I want to make sure I do my best at every meeting because there are a lot of people who planned to have that 15 minutes or 30 minutes or whatever.

And so I use Writer all day long to be able to really help with preparing for my day. For example, every time I put pen to paper and write something for Twitter or LinkedIn, I put it into Writer and I say, "Tell me what a technical person would take issue on this. How could this offend somebody? How could somebody troll me?" And, you know, it was really helpful. Even for me just to be able to take that and then write the rebuttals before I even post. Even just the act of writing out the rebuttals kind of gives me so much confidence to be like, “Yeah. I'm confident with this. Fuck it. Bring it on." And I think I write more, really, as a result of sort of having that confidence. So yeah, I use it daily.

I'm curious about how you want someone to feel when they engage with all this stuff, or when they use a Writer-powered experience. Like what do you want that person to feel when they use Writer to do their work? 

Yeah, it’s a great question. I think what I want folks to feel is powerful. Things that would have taken months or years for them to do, they can do in weeks. They can do it in a quarter. You don't have to wait for next year to do something. You can do it now. What would you do if talent and intelligence were unlimited on your team? What would you build? Those are some of the things that are possible now. And you know, the real kind of power in that is something that I think our power users really, really love. 
 

I wonder if you can dig into that “what would you do?” question a little bit more. Because this is something that we think about a lot—this idea that if you have a tool, whether it be Writer or something else, that helps to save you time, helps you work more efficiently, what can you do with all of that time and energy that you save, that you have left?

So much of the insight and the creativity that leads to the, like, how to spend your time actually comes from talking to your customers, talking to your audience, talking to your market, getting inspiration. And sometimes—we've all been there—the work gets in the way of the real work! [Laughs.] So, you know, that's what we want to free up for people. And honestly, that's what Writer has done for me. I have been so much more focused in each of my meetings. Being so adequately prepped versus thinking about my next thing, versus Slacking somebody while I'm with somebody else. It's horrible.
 

[Laughs.]

This is the part—like, looking into the whites of somebody's eyes is just the most fun part of, not just being at work, but being human. And that's what I hope AI—that's the best version of what AI can do, and we're working towards that. 
 

What are you looking forward to next? Both for Writer, but also for the adoption of AI powered tools?

Being able to really string together workflows—creating agents that help us, not just do task completion, but really work orchestration—I'm really excited for that. I think combining that with being able to do it with our minds is what has me very excited. Don't laugh. I tell my co-founder all the time, like, “I want to be able to go through my email in the shower while thinking about it,” you know? Or on the playground, right? I think it was Sam Jacobs—I don't know if you know him, he's great—he wrote recently: I want AI to mean that I'm at more dinner parties. I think I will also still be working at the dinner party, but there's a lot coming and soon, like three to five years time frame, that is going to render work unrecognizable. 
 

May, it's been really nice chatting with you. Thank you so much for joining us today.

Likewise. So fun, Matt, to go deep with you. Thank you.
 

~ ~ ~
 

A lot of AI tools are solo, singular experiences. But writing is a team sport. That means we need tools that don’t just help us, but help us work more effectively with others, too. Collaborative tools that understand how teams work, how they think, and what success looks like.

Because, like May says—if AI can help you spend more time in flow, that’s more time left for your customers or your team. Or for doing the work you love most.

To learn more about Writer, visit Writer dot com

Working Smarter is brought to you by Dropbox. We make AI-powered tools that help knowledge workers get things done, no matter how or where they work. 

You can listen to more episodes on Apple Podcasts, YouTube Music, Spotify, or wherever you get your podcasts. And you can also find more interviews on our website, workingsmarter.ai

This show would not be possible without the talented team at Cosmic Standard: Our producers Samiah Adams and Aja Simpson, our technical director Jacob Winik, and our executive producer Eliza Smith. 

At Dropbox, special thanks to Benjy Baptiste for production assistance and our illustrators Fanny Luor and Justin Tran. 

Our theme song was created by Doug Stuart. 

And I’m your host, Matthew Braga. Thanks for listening.

~ ~ ~
 

This transcript has been lightly edited for clarity.