How Writer is helping Fortune 500 companies become AI-forward

Note: Transcripts are automatically generated from episode audio, and are not fully corrected for spelling, grammar, and formatting.

MAY HABIB: You know one thing that AI never asks is why did you ask me that question? Right? AI’s always like, oh, I’m so glad you asked, or you are brilliant, right? Like, thank you for asking me that. Let me tell you blah. Right? It is incredibly subservient.

RANA EL KALIOUBY: That’s May Habib – CEO and co-founder of the AI studio, Writer.

HABIB: And I think the meaning of being human is to not be right. At Writer, we’ve got this saying, break rank, break glass, and that’s what I think it means to be human compared to being AI.

EL KALIOUBY: May and the team behind Writer take no issue breaking rank in Silicon Valley. They’re shaking up the status quo with an army of AI agents that can handle all sorts of enterprise tasks that humans normally do – from writing website blurbs to drafting business proposals.

These are often time consuming, menial tasks – and automating them is saving companies millions – companies like Salesforce and Uber, to name a few on their client roster.

Writer, which May founded in 2020, is now valued at 1.9 billion dollars.

On this week of Pioneers of AI, May and I are digging into the nuts and bolts behind one of Silicon Valley’s rising disruptors. We’ll talk about human-AI collaborations, data privacy, and building a unicorn start-up in the age of AI.

I’m Rana el Kaliouby and this is Pioneers of AI – a podcast taking you behind-the-scenes of the AI revolution.

[THEME MUSIC]

May, welcome to Pioneers of AI. It’s so great to have you on the show.

HABIB: Thank you Rena. It’s so nice to be here with you.

EL KALIOUBY: Yeah, so we have a lot that we share in common. We’re both of Arab descent. You, I believe, grew up in, were born in Lebanon and then grew up in Canada. Is that right?

HABIB: Yeah, that’s exactly right. We went back and forth a lot. So I definitely feel very Lebanese.

EL KALIOUBY: Uhhuh? Do you still go back?

HABIB: I haven’t been in a while, since before COVID. You always say next year. But Inshallah, next year.

EL KALIOUBY: Inshallah. That’s great. We’re also both young global leaders at the World Economic Forum.

HABIB: It’s such a great community.

EL KALIOUBY: It is. Yeah. My most burning question for today actually, is that you and my daughter Jenna both graduated with a major in near Eastern languages and civilizations.

HABIB: Yes indeed. Yes.

EL KALIOUBY: But somehow you ended up in tech. So I feel like there’s hope for my daughter.

HABIB: Yeah. It’s gonna be the only industry left. It’s so exciting.

Copy LinkFrom language tools to an enterprise AI platform

EL KALIOUBY: Yeah. Amazing. After years of working in venture capital, May founded her first company, Qordoba in 2015. It started out as a software company, helping their clients write clear, consistent content across languages.

HABIB: Yeah, it was a machine translation company. We did everything from human translation to machine generated translation, all in a platform that helped enterprises really build highly localized products and doing it in an engineering friendly way.

EL KALIOUBY: Qordoba allowed companies to enter new markets by translating all of their content into new languages with ease. But when transformers hit the scene – the kind of neural network architecture that bred LLMs – May saw new opportunities.

Her work didn’t need to stop at translation – generative AI opened the door for all kinds of business automation. In 2020, she co-founded Writer.

So you describe Writer as a full stack AI powered platform that helps customers, your enterprise customers, become AI first and agentic first. Unpack that for us. What does it actually mean?

HABIB: Yeah. So in 2023 we said dominant design in generative AI looked like LLMs and retrieval and guardrails, and an AI studio. An AI studio, because you really needed business and subject matter experts to collaborate.

And we called it a dominant design architecture because we said this is the only way to build highly reliable, effective AI systems on top of highly non-deterministic technology. And two years later everybody has an AI studio. Everybody has an agent builder but compared to Writer, they’re still very thin wrappers on top of the LLMs.

And what we’ve been able to do year over year is really invest in those AI native primitives that make it really easy to actually connect LLMs to data, to workflow, to guardrails to produce highly precise use cases, insights, content in highly regulated environments. So we’re doing client onboarding, we are doing KYC, we are doing orchestrated digital marketing. Yes. And it’s agents that just work in a way that is just not happening on any other platform.

EL KALIOUBY: Basically, May is saying that Writer offers a lot of capabilities to their clients – whether that’s integrating their AI agents into marketing strategies or using them to help automate fraud detection. She says that her clients are seeing results.

HABIB: So it’s been super incredible to see our customers that we’ve had really lead their market. Customers like Uber and Salesforce and Accenture and Nvidia, who are using Writer internally to really streamline all sorts of operations.

Copy LinkHow AI agents automate real business workflows

EL KALIOUBY: So May, can you walk us through some examples of how your customers are using Writer in their workflows?

HABIB: Yeah, the agentic use cases are just incredible. At Uber, we’ve got an incredible support team there that is using our agentic AI in their documentation and support processes. So they use Writer agents to classify Jira and Slack requests. They assess urgency and they generate updated responses and content, right? Based on that. The AI agent is staging the updates in the CMS and they’re publishing it upon approval. And so there’s this very artful orchestration between the deterministic and non-deterministic aspects of this workflow.

And it’s incredibly amplifying to what the human teams are doing. Tons of manual handoffs reduced, reducing content maintenance and production costs, et cetera.

RFPs and business proposals are also incredibly rich for agentic AI. At a company called Commvault, a cybersecurity company, AI agents from Writer were automatically pulling RFP files from Salesforce. We’re associating them with the right opportunity.

We’re setting up a dedicated Microsoft Teams channel where teams can actually collaborate on the AI generated responses that are based on previous RFPs that the teams have been successful in, and then once they’re verified by people right there in the channel, we are automatically producing the final PDF of that proposal or that RFP and then sending it to the client.

Copy LinkWhy managing agents is more like managing employees

EL KALIOUBY: This is really interesting, with my investor hat on, one of the areas that I’m very excited to be investing in is this idea of an AI employee, right? Like an agentic AI, which is basically embedded into a workflow and is able to start with a number of tasks, but increase its ability and its scope of responsibility over time as it learns and evolves. But it’s working alongside human teams. And your example of working on the RFP proposal – I have this picture in my mind. There’s the Microsoft Teams channel with all these humans and the agentic AI. Is this kind of how it works?

HABIB: So I don’t like to anthropomorphize AI. But the mental model of managing agents like employees is actually more apt than managing agents like software, right?

But you actually do need an agentic PM that can think about how you get these agents to do things reliably, or head of product. Matan likes to say yes, agents don’t follow rules. But you can get them to behave. Really taking the build process from this very deterministic, software based approach to how you actually align behavior of an agent system. Using the mental model of an employee is actually really helpful.

Like in Writer, we build blueprints. We don’t build workflows because that’s very deterministic and that’s not what we’re really doing here when we orchestrate these agents against a goal, right? These are really goal oriented and can take lots of different paths to achieve a goal within guardrails.

But your blueprint for that agent system is kind of like your job description, right? For an employee. Your prompt setup, your grounding – that’s like how you onboard an employee. Your escalation paths, your fallback agents, those are like your managers. How you do performance reviews, how you do feedback and coaching, that’s your outcome tracking, that’s your evaluation setup.

That’s your retraining. And then we’ve also built just lifecycle management for agents. So how do you deprecate them if they cost too much relative to the value, or folks aren’t using them anymore, or they’ve been replaced by something else really —

EL KALIOUBY: — Like firing.

HABIB: Yeah, exactly. Termination, right? And so the mental model of them as employees. I think we can think of them as a super employee, right? An employee that really needs to be a member of multiple teams to be able to get their objective achieved.

EL KALIOUBY: We’re going to take a short break. When we come back, how Writer is increasing enterprise visibility in a post-SEO world. And the new job opportunities that agentic AI is creating. Stay with us.

[AD BREAK]

Copy LinkHow enterprises onboard AI and adapt to GEO

EL KALIOUBY: When you onboard a customer, what does this look like?

HABIB: Yeah. So we are actually a verticalized platform. And so when we onboard a CPG customer, it looks a little bit different than when we onboard somebody in retail, or onboard a technology customer. We have libraries of hundreds of prebuilt agents, and they’re quite powerful. So a Qualcomm that does really sophisticated ABM – Account based marketing, or a Salesforce that does incredibly powerful comms and PR, right? Their use cases are going to be different. And what we’re able to do is onboard them into libraries of prebuilt agents that are ready for configuration for their systems, right? Their tooling. If you are doing GEO, that’s generative AI engine optimization, rewriting –

EL KALIOUBY: Actually, by the way.

HABIB: It’s a new term. Yeah. GEO is the new SEO. When you really think about how much of the consumer market is turning to LLMs for search, right? Really thinking about your web footprint as a brand and what it’s going to mean for the visibility of your products, your SKUs, your viewpoints. You absolutely can generatively restructure all of your site map to be much more amenable to the kind of web search patterns and the training data collection of LLMs.

And I’ll give you like the tiniest tactical thing that we do. So much of conventional search is keywords based, but LLMs really like nice comprehensive paragraphs and sentences in their responses. And so actually being able to mirror your content automatically, right? It’s exciting to get to take solutions to customers where generative AI is a solution to a problem that generative AI caused. But it really gets to the change management complexity.

EL KALIOUBY: Yeah. So I kind of wanna simplify this transition from search engine optimization, which is essentially search based on keywords, which we’ve had for the last, I don’t know, like 20, 25 years, to this transition to generative AI engine optimization – and I love that term – which is essentially modifying or adapting your content so that it could be found by a gen AI search engine, like Perplexity or whatever’s out there. That’s super cool and I think that’s going to also create new business models and monetization, right?

HABIB: Yeah, absolutely. I think the consumer LLMs are likely to do the kinds of things that Snapchat did, right? Like brand takeovers, exclusivity in certain categories. So there’s definitely going to be an explosion of that. And I think the brands that really get there first will have first mover advantage.

Copy LinkWhat it takes to become an AI first company

EL KALIOUBY: So a lot of the work you do with your customers isn’t just about the technology, to your point, it’s about change management and bringing the organization and the employees on board with all of these new tools and technologies. I imagine a lot of our listeners are grappling with how to do that, how to become AI first, how to incorporate AI into their workflows. What have you seen that works? Like what are some of the best practices that your customers are using?

HABIB: Yeah, so too many customers are saying help us become an AI first company. In fact, there’s a mandate to be an AI first company and they don’t back it up with the proper resourcing. Right? And we’ve got an incredible delivery team that works really closely – no daylight between us – with our customers in helping them, hands-on keyboard, really build out solutions that can go live in a matter of days.

Matter of weeks, but the real scale of the program is going to depend on, do you have agentic product managers that can rewire workflows that have been built up over decades? Do you have AI builders who – we can now address a whole range of software capabilities. It’s like vibe coding on steroids at Writer, right?

I mean, you’re literally writing a prompt of a process and we are unfurling the identified version of that, right? That you just really need to modify – the code is there already – that you’re able to go out and modify. And so when you think about what that means, we have made building tooling and software a thousand times easier in the enterprise. They can go straight to build it and scale it because the guardrails are really powerful.

But back to your question, this is a lot for organizations to think about. We are seeing best practices be, yes, I’ve invested in full-time folks focused on generative AI, and I think about it as two in a box down every team, engineering and the business together, because you as a technology team aren’t gonna be able to build agents for the enterprise in an isolated way. You need them there for a highly iterative collaboration. And that’s just a different set of skills that we really help mirror for our customers, in the people that they have put up to say, yes, this is our agent PM, this is our agent builder. These are the people who are gonna build on Writer.

EL KALIOUBY: So agent product manager and agent builder. These are new types of jobs. Can you talk about what these look like?

HABIB: Yeah, absolutely. Your product manager, right? Your junior product manager is gonna be somebody who can absolutely transition to be an agent PM. But they’ve got to start thinking about the agent development lifecycle, right? Versus software development lifecycle. And remember what we said at the beginning – agents are systems that are goal oriented, that can create their own steps and workflows, that can rebuild those steps as they try to achieve a goal. And so the requirements gathering is no longer, oh, legal needs a chatbot. It’s alright, I’m gonna build an agentic system that cuts down my contract review time. Right? And that kind of outcome orientation, and the specificity of the problem that we are solving – much more than here’s what the UI should look like, or here is what the steps should be to solve this problem.

That’s a change in mindset. And so yes, it’s a process design. That’s an input for things that are existing processes inside of a company. How you pay an invoice or how you onboard a vendor in your supply chain, or how you onboard a customer into your mortgage product.

But it is the behavior design of the agent that you are building to address those problems that we really want the PM to own.

Copy LinkWhy speed culture and customer feedback are the real moat

EL KALIOUBY: What differentiates Writer and what’s your competitive moat?

HABIB: Yeah. This is a space that moves so fast. Speed and iteration is the only moat. And I think you can boil that down even further. Culture is a moat in this space. And when we look at our business, we’re very proud that our top 30 ish customers are at 217% NRR. This is customers that are tripling spend with us.

Global 2000 customers in the trailing 12 month period. And it’s only going up. And a huge part of that is the multi-year investment in the architecture blocks in a highly disciplined way, maniacally focused on the enterprise, and it starts with our AI research team. We got to synthetic data before anybody else.

We got to graph based retrieval before anybody else. We got to self evolving models before anybody else because we’ve got this really exceptionally fast feedback loop between the customer and the research team, right? So it is really powerful to be able to take problems that we’ve got in market to a team that is right there at the table with us.

That literally plans a research roadmap around how we get this level of adoption and explosive functionality in super intelligence in the hands of the companies that we work with. And the partners that we work with. You have enterprise AI too, and I think there will be a lot of winners.

Your traditional software companies are going to spend a lot of time building enterprise interfaces into their own data and into their own products. And it’s a really big market for the workflows that don’t exist because it’s too hard to create these interactions between data silos, right? And the unstructured data that sits in people’s heads that really define how to do something. And so we play that role in a really unique way.

EL KALIOUBY: Unlike a slew of other agentic AI companies, Writer is not using off-the-shelf LLMs. They’ve built their own models and we’ll get to why they made this decision after a short break.

[AD BREAK]

Copy LinkWhy Writer built its own models for enterprise use

EL KALIOUBY: So we love on this show to take our listeners behind the scenes and kind of unpack what is behind the products you’re building. You mentioned a few key features of your products that give you a competitive moat. So let’s talk about those. First of all, let’s talk about your LLMs. I think it’s really interesting that you decided to build your own as opposed to build on top of other foundation models. How did you make that decision and why?

HABIB: Yeah. We were born as a native transformer company. The story of Writer is the story of the transformer. And our Palmyra X Four and X Five are the workhorses of agentic workloads. So token context windows in the millions of tokens. Breakthrough speed, the ability to process thousands of pages in 20 seconds. The ability to tool call in a microsecond. Adaptive hybrid reasoning. And when you think about the way that you daisy chain so many AI components together when you do agentic AI, it’s incredibly important to have models that are fit for purpose.

These are not small models. They’re 500 billion plus parameter models. But we have architected the transformers to be highly efficient. So the total cost of ownership is incredibly attractive to customers when they really understand that retrieval in agents and the platform are really all in one. The reliability, the consistency.

The last point is really around transparency. We give our customers more transparency into how we build our models, how we train our models than they get from open source, right? And that makes a huge difference in folks’ abilities to get through model risk. We’ve got a lot of customers in financial services and their ability to really have models that they trust under the hood. We get to help them with data that they consider highly confidential and really access it in a way that they understand is very safe.

We don’t use their data for training. The models are actually trained on completely synthetic data. But we give them the visibility they need into training data, weights, architecture, the guardrails that we use, the audits to limit bias and toxicity. You get to audit the model’s thought process. There’s explainability built in.

EL KALIOUBY: That is very unusual by the way. That’s very differentiating. Because most products out there, you’re using the LLM as a black box and you have no idea what’s going into it and also what it’s learning, right.

HABIB: A hundred percent. Now we are an agentic orchestrator, so you could use an agent to call out to any other model. We’re multi-model in that regard. Folks use Gemini for image generation. Firefly is a big partner of ours, but 95% of the workloads are on the Palmyra model.

EL KALIOUBY: So the Palmyra X Five you announced is able to get to state-of-the-art performance basically three times faster than the other models, and four times cheaper. And one of the key components, or secret sauces I guess, to this is your use of synthetic data.

So back in my Affectiva days, we would use synthetic data generators. For example, if we’re training a model to detect driver drowsiness, right? The traditional way is to look for examples and in fact go out and collect data of people falling asleep at the wheel, which is really expensive because you want diverse human beings and diverse situations and whatnot. So instead we had a synthetic data generator that generated synthetic humans and we could change the parameters, change the way they’re looking, change the degree of drowsiness.

So walk us through how are you generating the synthetic data? What’s the input to the synthetic data generation models?

HABIB: Yeah, well it’s secret sauce. I won’t tell you too much, right. From a data pipeline perspective. What I can share is think about our use of synthetic data as the creation of training data that is precision created for specifically the training of models. And so the way that we structure the data that’s the input to a model that we fine tune to actually synthesize data that is fit for purpose, for training the models that we want to have certain types of behaviors. And so we align our models based on what we want them to do, and the synthetic data really flows from that.

Copy LinkHow synthetic data and privacy shape trustworthy AI

EL KALIOUBY: Actually this is a good segue to what I find really interesting about how you are training your LLMs, which is that you’re not actually using your client’s data and you have a zero data retention approach. Tell us more about that. Because in the world of machine learning, data’s everything and everybody’s like trying to capture every little bit of data, but you have a very responsible approach to how you use your customer’s data.

HABIB: Yeah, absolutely. I think it’s become table stakes to not train models on folks’ data, right? OpenAI’s not doing that either anymore, after a certain level of spend. But we don’t have to retain the data. There’s absolutely feedback that users are able to give us manually from within the product and certain telemetry signals that we are able to infer from. But we can be really generous in how we do the data retention. And it’s actually so modifiable that you can just change it in our UI, in the IT command center.

Copy LinkWhy AI adoption works best when employees benefit too

EL KALIOUBY: Very cool. Okay, so let’s talk about AI replacing jobs and helping companies scale with fewer employees. You have a different perspective that you shared at Davos. You believe that companies that rush towards an AI led future without bringing their employees on board will be at a disadvantage. Say more.

HABIB: Yeah, I think it’s like getting turkeys to vote for Thanksgiving. It makes no sense for me to learn this technology, bring it to my company if I know I’m doing it to get axed, right? Like, am I really gonna download my brain into this agent so that you don’t need me anymore? Right?

And the reality is, yes, to do this body of work now, if I were building it AI first, I only need a fraction of the people that are here, but it doesn’t mean that’s how I’m going to actually operate or run my business, right? We’re gonna launch new products, we are going to enter new markets. We are going to go after new segments, and that is a much more empowering and exciting message from executives that helps them win, right? And yes, will there be less need to backfill attrited employees, or headcount avoidance that comes from becoming AI first? Sure. But I don’t see anybody – nor do we advise it – going about and axing people because AI is coming. The most successful customers that we have seen are going to groups of 300, 400, 500 in healthcare – nurses, doctors, clinicians – and saying, look, we’re gonna take on another customer. We’re gonna empower everybody with AI and we’re gonna double our business. And it’s actually happening. And that’s the success story.

EL KALIOUBY: Yeah. I love that you also shared that AI systems will only be as strong as the diverse perspectives that go into developing and deploying these AI technologies. This is something I’m very passionate about, as you’d imagine. Why do you think so? Why do you think we need these diverse perspectives?

HABIB: Yeah. This is technology that is going to make a lot of companies much more powerful and a lot of people much more powerful. And we’ve got the potential to build a much more equitable society as a result of the technology. At the same time, everybody’s got a teacher in their pocket, a doctor in their pocket, right?

I’m literally using AI all day on the weekend because I’m learning new stuff. I’m a healthcare freak. And so I know way too much about HRV, thanks to AI, and that doesn’t cost any money now, right. And I think really being able to build that technology in a way where everybody feels represented, right?

If I’m asking a health question and the answers are clearly from a point of view that doesn’t represent me, then I’m not going to feel as comfortable in that product. And I think especially in consumer facing tech, folks who really think about who their audience is and think about building LLMs accordingly are just gonna be more successful.

EL KALIOUBY: Yeah, absolutely. And I would be remiss to not mention that you are one of the very few women who are a CEO of an AI company that’s a unicorn. So I just love that and congratulations.

HABIB: Thanks, Ana. There are a lot of amazing women at Writer. We’re more than 40% women, and pretty even across all teams. So if you’re listening and you are of any and all genders, please message me. We are hiring like crazy across absolutely every function. We’re hiring a lot of industry domain experts as well. So if you come from a world where you’ve got specialized expertise and want to bring it to enterprises trying to transform their own businesses AI first, please get in touch with us.

Copy LinkWhy diverse teams will build better AI products

EL KALIOUBY: I love that. I love that. What keeps you up at night?

HABIB: I mean, I track my sleep pretty religiously, so not that much.

EL KALIOUBY: An Oura ring or a Whoop? I’m a Whoop.

HABIB: I’m an Oura.

EL KALIOUBY: Yeah. I feel like Bay Area’s Oura.

HABIB: Yeah, it’s won over the West coast. Yeah. But this is an incredibly fast-paced space. I got to my desk at 4:00 AM this morning because I was so excited for my day. So there’s a lot. Hard work and people — well, I fell asleep at eight, so Oura was happy. I mean, I shoot for a score over 80, so. I was like 81 or 82 or something.

EL KALIOUBY: Fine. Fine. I love that. Thank you for joining us today.

HABIB: Thank you, Anna.

EL KALIOUBY: Talking to May feels like a glimpse into the future of work. A future where AI agents will be embedded in all types of businesses, doing all kinds of tasks. One where executives use this superpower as a mechanism to expand their markets, while building their human talent pool.

But what I find so striking about companies like Writer, is the organizational shift they offer. Imagine a company org chart that lists AI agents, like the ones Writer offers, working alongside humans. While I agree with May that we shouldn’t anthropomorphize AI, I do think we’re approaching a new working order. As an investor in this space, I personally am really excited about these kinds of restructures and its implications on the future of work. I do think some jobs will disappear, but I also agree with May that AI will create new job opportunities for humans.