Skip to main content
Abstract image representing AI technology.
What if? So what?

What if AI Was Your Most Strategic Hire? An Interview with May Habib, CEO of Writer.

In this episode, host Jim Hertzfeld engages in a captivating discussion with May Habib, CEO of Writer, about the transformative potential of generative AI. From exploring the evolution of AI to uncovering strategies for overcoming adoption barriers, May shares invaluable insights that will help listeners navigate the dynamic landscape of artificial intelligence.

Connect with May Habib, CEO and co-founder of Writer

Connect with our Host, Jim Hertzfeld, Area Vice President, Strategy

Subscribe and Don’t Miss an Episode

Listen on Overcast

Meet the Guest

May Habib

May Habib is CEO and co-founder of Writer, the only fully integrated generative AI platform built for enterprises. Leading brands choose Writer to help them automate key operational activities and increase employee creativity and productivity. The company's security-first approach means that Writer proprietary LLMs are deployed inside an enterprise's own computing infrastructure. May is an expert in AI-driven language generation, AI-related organizational change, and the evolving ways we use language online. She's received many recognitions, including the 2023 Forbes AI 50 and Inc.'s 2023 Female Founder Award. May graduated from Harvard University and spends her time between San Francisco and London.
Special thanks to our Perficient colleagues JD Norman and Rick Bauer for providing the music for today’s show.

Episode 43: What if AI Was Your Most Strategic Hire? An Interview with May Habib, CEO of Writer. - Transcript

May (00:00):

The risk associated with trying out a tool is just really nonexistent. People expect you to now; it's almost gone the other way. It's sort of risky if you aren't proposing to your boss that you're going to use AI instead of hire somebody.

Jim (00:19):

Welcome to What If? So What?, the podcast where we explore what's possible with digital and discover how to make it real in your business. I'm your host, Jim Hertzfeld, and we get s**t done by asking digital leaders the right questions. What if? So what? And most importantly, now what?

Jim (00:34):

Hey everyone. This is Jim, and we're here with May Habib from Writer. Hey, how are you doing, May?

May (00:41):

Oh, great to see you again, Jim. I know it was only the summer, but it feels like years already.

Jim (00:45):

We're moving fast in all facets of our lives. So, before we jump in just a little bit for the folks who don't know about Writer, tell us a little bit about your company and what you do there.

May (00:55):

Yeah, happy to. Writer is a full-stack generative AI company. So what we mean by that is, it is the large language model, state-of-the-art large language model, and all of the tools around it that enterprises need to build useful custom applications. So, AI guardrails that are built in and reusable across use cases.

May (01:17):

A part of the product called knowledge graph, which connects to your structured and unstructured data, an application studio, which allows your teams to build their own tools. And so, what we've done is really say, look, what folks want is the value of generative AI. They don't necessarily want to build every layer of this stack.

May (01:37):

And so, we've put that all in one enterprise-ready platform that can be hosted inside of a customer's own private cloud. And so your data never leaves your environment. The model is there as well, and it's just proven to be a much faster time to value in generative AI, a much lower total cost of ownership, and end users are really happy, which is so important.

Jim (02:02):

I love that the concept of it, and you guys have been at this for a while, right? I think since 2020. But really, I would say sort of entered the public consciousness about a year ago. So, in this generative AI, what have you really learned in this last 12 months through this whole experience and this kind of emergent type cycle?

May (02:22):

Oh my goodness. It's been really an incredible amount of learning, I would say. Every year since we started, it's just getting faster and faster. And I, I do think that is something that we say a lot, the pace of change is the disruption, get used to it, get used to every year, feeling like you learned more than the previous two or three years, there is just a real excitement around-

May (02:43):

-so much of just fundamental AI literacy being much more prevalent. Like if you told me even a year ago that, you know, I would be having debates on my LinkedIn feed about retrieval augmented generation with both technical and non-technical digital leaders. I would tell you you're crazy, but that is what's happening.

May (03:01):

We can talk to CEOs of Fortune 500 companies who are familiar with the scaling challenges their companies have with vector search. So, everybody really wants to see a transformation happen with generative AI. The promise of more intelligent workplaces is so exciting. Our vision is to transform work and we do want to help companies do that.

May (03:25):

There's a lot to learn for all of us and, luckily a real excitement around making that happen, you know, from, from customers and from us, it's a very close partnership with the customer. So, what we learn, they learn what they learn, we learn, and that that's been part of the year as well.

Jim (03:41):

And I think you mentioned earlier that there's a comfort level or a risk acceptance, right? We've kind of broken through some of those barriers as well. Right?

May (03:49):

Yeah, a hundred percent; the risk associated with trying out AI tools is just really nonexistent; people expect you to. Now it's almost gone the other way. It's sort of risky if you aren't proposing to your boss that you're going to use AI instead of hire somebody. And so people really want to see-

May (04:08):

-that, especially if in your leadership position, you're thinking about how to leverage AI in your workflows. And it makes sense to me as well, right? There have been such incredible capability gains, right? You shouldn't be planning to do everything the same in ‘24 that you did in ‘23.

Jim (04:23):

That's a great observation. We went from, well, we can't do that to wait. we have to do that. So, we're seeing it even in our business. From your perspective, what do you think is the overarching reason that organizations need to embrace or leverage AI in their digital strategies?

May (04:37):

First of all, the applications in customer experience are probably the most impactful and transformational. Really, finally being able to do things like personalization and segmentation, we've sort of had the piping to do all of that for a while, but really no bodies to produce all of those various derivatives and variations that you need for segmentation and personalization to be.

May (05:02):

To be a reality. And now, with AI, we have seen that work incredibly well. From Salesforce to Accenture, we've got customers who are really kind of winning internally with programs that personalize and segment so much more than they were doing before. And the digital journey is probably also the one where-

May (05:22):

-there is the most amount of, reviews and workflows and handoffs between a writer, an editor, a program manager, a client, a brand specialist, a legal expert, a compliance expert, back to the proofer, back to the person who loads it up, there's a lot of work about work in the customer experience.

May (05:41):

And so, if you are a digital professional, the part of your job that probably brings you the most joy is insights and audience development and talking to your users and your customers and thinking about strategy. And it just ends up being not so much of your day or your week, and that goes for your teams.

May (06:00):

And so if you've got digital in your remit and you haven't taken a serious look, it's definitely something to think about.

Jim (06:08):

Yeah. You bring up a good point and just in general around automation and AI about how we can get to the fun parts of our jobs, right? If we can appeal to our strengths and take some of the noise out of it, you know, but I think despite some of these benefits, there are still some laggards out there.

Jim (06:21):

There are some organizations are still, are still hesitant to integrate generative AI into the workflows and into their business processes. What do you think is preventing organizations, especially the larger ones, from embracing these technologies?

May (06:36):

I think it's probably three things, Jim, I think number one is genericism. Yes, it's exciting to get your kid's party invitation written out and like, gosh, that was a really fun skit I wrote or poem I wrote with ChatGPT.

May (06:50):

But many folks who use it for work say, okay, it's kind of like remixed Wikipedia. I'm not really certain how I would use this, especially in digital roles, right, this is nowhere near the quality. It really is the number one reason I think reason folks aren't, leaning in is, is the quality they have seen from the consumer tools just doesn't meet their own bar. On number one, there are folks that'll say, well, you're not using it right.

May (07:10):

Ask it to critique what you just wrote and ask it to give it to you in different voices. But, like the core heart of the most impactful thing they do, could they see AI helping with that? A lot of folks, first blush, say no. Second, being able to really feel safe putting your data up there.

May (07:29):

Whether it is, getting the insights out of this report or compare this to that. Not only are folks personally wary, but a lot of enterprises have just shut down access, right? You cannot put company data into any of these tools where the folks who run these models own the data and all that comes with that.

May (07:47):

And then I think that the third category of risk that people associate with these tools is inaccuracy, is hallucination. That's a separate point from quality. If I do ask it this question, is it going to give me the right answer on my data? And there are a lot of reasons to really feel that-

May (08:05):

-actually, that is the riskiest one, number three. So those three things together, do add up a lot of risk. The very largest organizations in a lot of ways have had the most to gain and the most bodies to put against resourcing, governance and risk adjudication and security and legal and all of that.

May (08:25):

I do think it's actually smaller organizations that have said, okay, the risks are too high. And we aren't resourced right now to be able to go put, 10, 15 people here. And so, I think there are a lot of things that folks like us could do, folks like you could do and we are doing together to help educate folks about the risk reduction, right.

May (08:44):

And, and really maximizing building custom solutions that get around, you know, the genericism or lack of quality. And connect the application to own their own internal sources of data that give them much more accurate valuable results.

Jim (08:59):

When there's risk, there's opportunity. I know there's some, some wisdom there. Right. So, I think those are great angles. I think it's, it's interesting. You know, the vocabulary that you're hearing, from the business sides and then the customers that you're seeing are talking to you.

Jim (09:13):

These terms are sort of coming into the consciousness, right, as well. So, given some of these barriers, some of these concerns some of the things you guys have learned, what, what is Writer doing about it? How are you guys responding to this?

May (09:26):

Yeah, well, the first thing is we've built these reusable AI guardrail modules, so on fact-checking and accuracy, if we think there is a hallucination potential in an answer we gave you or a generation. We will highlight the facts, the figures, the quotes, the stats that we think probably don't have a basis for reason, but, the model is going to do what the model does and answer your question, and it's ungrounded in your data, what we just spit back at you.

May (10:01):

That's number one, number two, and those are reusable modules and there's an API for that. So, if you use Writer to build in a generative AI application, you can still use those features that we've built into our own UI. The other sets of features within AI guardrails that are reusable across generative AI applications have to do with brand and editorial and brand voice and style and all of the things that we work so hard to make sure-

May (10:28):

-are consistent across, you know, hundreds of people or thousands of people that might be writing for the customer experience. And now suddenly everybody's got an AI Writer sub-tool somewhere and you have just multiplied that problem. A lot of marketers we talk to say, volume is not my problem.

May (10:46):

I don't want more s**t. I want less s**t. I want better s**t. And, you know, I need it to be just much more effective, much more consistent, much more on brand, and those have been some of the AI guardrails that folks appreciate the most, whether you are answering sales, you've built a sales enablement app with Writer and, you're giving a salesperson the kill bullets or the objection handling, you're doing it using-

May (11:13):

-the terminology and the messaging that you have developed as a digital leader. So, brand guardrails are really important overall AI guardrails for accuracy, for risk, for toxicity, the governance AI governance and AI task forces and security that has really spun up around governing generative AI.

May (11:33):

Folks have come up with their own guardrails for how they want people using it. And so, they need to be able to plug in. You can't be kind of rebuilding every generative AI application and then needing to screen for guardrails. And so, we've built that right into the product.

Jim (11:47):

So, you just say really facing some of these things head-on, right? I mean, that's, that's kind of what I'm hearing.

May (11:51):

Yeah, you know, it's, it's a SAS solution that doubles as AI infrastructure. So you're able to really get the utility and power of a state-of-the-art, large language model and a graph-based retrieval augmented generation solution sitting inside of your own cloud environment, but you don't have to manage all of the use cases and business users and last-mile data and annotation.

May (12:18):

There's a whole platform that does that for you.

Jim (12:21):

Yeah, that's cool.

May (12:22):

For you isn't the right word, with you is the right word, right, and not having to rebuild a lot of those pieces every single time you build an app.

Jim (12:30):

You know, that's, that's like, we're going to make this even bigger. That's advice I heard a long time ago. God doesn't work for you. God works with you.

Jim (12:36):

So that's...

May (12:37):

I like that. Ha ha ha.

Jim (12:38):

I hope this inspires them to, to make a move, to do something that they haven't done.

Jim (12:42):

What's one piece of advice or one step you would you would tell someone to kick off their own journey?

May (12:48):

Mm. Great idea. Someone who hasn't done anything yet.

Jim (12:50):


May (12:51):

Yeah. I selfishly want to say, hey everybody listening, call us we're Hello@Writer we can help. But whatever it is that's approved and in front of you and, and we have a free product and a free trial whether it's ChatGPT or BARD or really anything figure out if you can use it one or two times a day.

May (13:09):

Get the ChatGPT app on your phone and literally just like, how should I take care of this, bunny that I need to wash? It smells like vomit. Like, should I dry clean it? I mean, that's literally the last time I used ChatGPT. You know, I, I don't use Writer for that stuff, but like your consumery around the house things.

May (13:27):

ChatGPT does the job. Just figure out what it looks like, right, so that it's not really scary that you've turned it into, Google Maps and Uber and there's AI built into all of our products. So simple search, right? Finding answers versus resources using generative AI for that. And then at work, I think really reaching out to peers. So, we've got a great community of chief digital officers, chief marketing officers, very happy to connect you with folks who might be familiar with your industry. We do a lot in retail, financial services, healthcare, technology (not surprising), professional services, of course, and really speaking to folks who have done the rollouts, who have done the proofs of concepts before that who have really figured out how to manage that last mile user education.

May (14:15):

That's a whole other ballgame really make sure there's broad and equitable distribution of this tooling once you do introduce it, because everybody adopts at a different pace. A lot of these workflows change pretty dramatically, and you want to bring people along. You don't want that to be sudden.

Jim (14:32):

That's awesome. It sounds like what you're sharing is a lived experience. You guys have been at this for a while. I love the pragmatism. We love pragmatism here at Perficient. We love to be of a no-nonsense, getting in the weeds, get it done. I love the advice. So just start, right?

Jim (14:46):

Just start. You want to finish. You got to start so, great advice. Thanks so much for taking the time to be on the show, and for the rest of our listeners, until next time, keep asking what if? So what? Most importantly, now what?

May (14:59):

Mm. Love it.

Jim (15:01):


May (15:01):

Thank you, Jim.

Joe (15:02):

You've been listening to What If? So What?, a digital strategy podcast from Perficient with Jim Hertzfeld. We want to thank our Perficient colleagues J. D. Norman and Rick Bauer for our music. Subscribe to the podcast, and don't miss a single episode. You can find this season, along with show notes, at Thanks for listening.