He was so tired that he crawled right into bed when he arrived home. But damn, he was happy.
He smiled and remembered the day. A rooftop full of bright and smiley people.
Artists creating artwork, wonderful conversations, part-time DJs, and even an amazing recreation of Grady the G.O.A.T.
A lot of work is required to make a big event like this one see the light of day, and it was worth it.
He thought back to a couple of years ago, when he spent weeks and weekends brainstorming at a whiteboard with his co-founders about what this company would solve.
Now, they were growing faster than ever, throwing amazing parties, like this one on a rooftop at sunset in Seattle.
But they had to build it from scratch, piece by piece, and rebuild it for the 10th time as AI models kept shifting under their feet.
Which wasn't exactly a problem. His background never gave him a map, either.
Just the confidence to work without one.
Missed Part I? Read the full article on our website.
Welcome back to Zero to Unicorn, Unicorner’s bi-weekly deep dive into life stories and lessons of the founders behind exceptional companies.
In other unrelated news… Tech Week is around the corner. We have many events coming your way in both SF and LA. Just to get a sense of who we can expect, give us a light RSVP so we can clear the room!
Now, presenting Part II of our deep dive on Gradial.
Picking up where we left off, Doug Tallmadge, our “why-asker” physics bachelor who loves tech, energy, and AI, had left his job working on Starlink at SpaceX to build his own company.
What this company would solve when he decided to become an entrepreneur was still a mystery. The main idea was using AI for enterprise, something not obvious back then.
After all, people were (rightfully) excited by AI chatbots, while Doug, always thinking about problems that don’t exist yet, was already pondering the future.
The first weeks trying to organize this newborn company were pure joy and excitement… followed quickly by a “What have I done?”
So they experimented, and found a fit with marketers. Marketers are scrappy, usually running lean with small teams and tight budgets, which pushed them to be among the first to try AI to make the chaotic workday easier.
Saving money and doing more with less? We’re 100% down.
How did Gradial help? The product that stuck was an AI agent on top of all the CRMs, ERPs, CMSs, and the rest of the three-letter-acronym soup of enterprise tools a marketer has to deal with.
See, initially, Gradial's goal was to make content management less painful. That’s because, yes, creating content became easier with time thanks to generative AI, but the management part was an awful experience, pardon our French.
Gradial applied the latest generative image and text models to existing design systems so content and digital experience teams could design websites, build multi-channel campaigns, and make content updates faster by asking in English. In other words, removing some of the maintenance and management overhead marketers deal with on a daily basis.
Soon, the early concept of an AI agent started to take shape. It was vague. Remember, the agentic AI trend we see today was not yet a buzzword.
But perhaps, there was a future where marketers would have multiple AI agents helping them do the boring parts of the job.
"This execution gap was the first problem that we sought to solve at Gradial, which is, ‘How do we put agents into this workflow to help do some of that manual work and actually help that workflow execute?’"
Nobody on the team had ever built something like that before, so they built it from scratch.
Over, and over, and over again.
They rebuilt their technical harness somewhere between 10 and 20 times as the models got more capable.
There was no textbook for this. There was no step-by-step recipe to follow to build an AI agent back then… But, it worked.
Some of the members of Gradial’s team at the holiday party in Seattle, 2024. (Gradial’s LinkedIn)
This was the idea of "a true digital coworker that helps you launch more campaigns, move faster, and scale effectively.” Doug’s physics background helped, since he was used to working through problems where he didn't know the outcomes to ahead of time.
Soon, Gradial would receive $5.4 million in seed funding in February 2024, led by Madrona, with participation from General Advance, Outsiders Fund, and Space Capital.
In addition, Gradial reported 30x YoY revenue growth in 2024 and grew fast enough to need two office upgrades.
Sometimes you're so fast that you bend space-time and arrive in the future while others are still figuring out the present.
Unsurprisingly, this would attract a ton of attention.
This seed round was the first piece of a domino effect, because just one year later, a Series A followed.
In March 2025, Gradial received $13 million in funding, led again by Madrona, with participation from Pruven Capital, General Advance, Outsiders Fund, and DLA Piper.
At this point, Gradial had reached what Doug calls “critical mass” of customer references. This was the flywheel for go-to-market, which some may say matters more than the product itself.
Essentially, as Gradial saw successes, it began to attract more and bigger brands, because better references in turn led to better demos and domain fluency.
The product itself evolved a lot, too.
One of the first things the company found success with was reliable JSON generation, structured output that at the time was inconsistent from frontier models. So, it sold JSON generation as a service.
It worked well for a while, but the models soon caught up. Since it wasn’t a good product differentiator anymore, the team needed to adapt fast.
But Doug wasn’t surprised.
“This is probably something that the model providers are just going to do, so we didn’t build our company around it… You always need to be adding more in software.”
At the same time, it pushed the team to continue focusing on product specialization.
Tools that try to serve everyone end up serving no one particularly well, while going deep in one domain means understanding its workflows well enough to resolve obscure (but meaningful) pain points.
“It is very hard to know the ergonomics of the workflow if you've never worked in the workflow.”
So, Gradial kept specializing in marketing and marketers' daily activities, instead of trying to be a horizontal, industry-agnostic tool. If the team could know these workflows like the back of their hands, they could create the most valuable product for marketers every day.
When you understand that, it’s clear that if done incorrectly, another B2B tool just means more overhead, not less.
"One of the things we've prioritized is working where our customers work. We don't want to give them yet another tool or platform they have to put stuff into. The question we're trying to solve every day is how we best fit into the workflow of the customer and help them do their job more efficiently."
It's no coincidence then that one of the Gradial’s most recent slogans is “Take the work out of your workflow”.
Gradial’s team at its brand-new Seattle headquarters in 2025. (Gradial’s website)
Nine months after its Series A, in December 2025, Gradial raised its Series B. The Series B came at a valuation of $350 million, led by VMG Partners, alongside Madrona and Pruven Capital.
That’s fast. We can’t emphasize that enough.
Create a product so irresistible to one specific group and its pain points that they can't live without you. These companies will spread the word and build success cases for you.
In Doug's perspective, if people aren't banging down your door after a Series A, ask yourself why not. If the issue is visibility, that's when you lean on awareness-building, like events and LinkedIn, to fix it.
That visibility brings interest. One investor emailed him seven times before getting a response. But that investor ended up becoming a key partner.
Being deliberate with his investors is intentional.
Remember Tyler from beehiiv? He too selected investors that were aligned with his line of business and interest. In his case, it was long-form content creators.
So, saying Gradial was doing “pretty well” would be an understatement.
But somehow, the momentum continued, faster and faster, blowing through the roof.
After solving the operational layer for marketers, making it easier and faster to connect common software used by marketing enterprise teams, Doug felt something was… missing.
"It's not enough to just have agents in the workflow if you don't have context. And this context layer is increasingly stuck in the existing MarTech systems that just weren't designed to power agents. And so, you know, this is something that we've been excited about over the last three years and really building towards at Gradial. But we're finally at a point where we actually have agents that can not only execute on the workflow but also this opportunity to really reinvent the infrastructure underneath to support agents as well."
Thus, Gradial had a massive launch in June 2026 alongside its Series C funding.
The official launch of the Agentic Content Infrastructure in June 2026. (Gradial’s launch video)
That’s just six months, for those keeping track.
With it, Gradial introduced an AI agent orchestrator at the center of Gradial’s platform, lovingly named Grady the G.O.A.T.
Gradial at the Forrester CX Forum West event in July 2026. (Gradial’s LinkedIn)
At this point, Gradial’s next step became building an agentic infrastructure underneath the workflow that belongs to your company’s cloud. The vision is a library of knowledge, the context the whole marketing of your company sits on, centralized in a single place. Think something like 20 different tools, all in one place.
“Agentic conference content infrastructure is taking all of the different pieces of your marketing context and bringing them back into your cloud. Bringing them out of all of the SaaS tools that they're trapped in into a single data lake that's sitting within your enterprise cloud that you control, where you're not having to pay the gas tax to access your data every time an agent needs to work with it. You're not having to go hit some SaaS API to access the data in a different system for every part of your workflow. And instead, over time, you're actually getting all of that enterprise context back into your cloud in a foundation that you can then build and scale agents on top of.”
That’s easier said than done. Legacy marketing tools were not designed for the speed agents demand, and this is becoming more and more evident.
Gradial did it, though. Alongside its $65 million in Series C funding, it boasts agents working at companies like AWS, T-Mobile, Vanguard, Kaiser Permanente, U.S. Bank, and Prudential.
Marketers need something that seamlessly integrates the tools they use daily and also all the context happening in between. Like one big brain that can answer what, who, when, and how.
It’s reminiscent of Doug’s childhood curiosity.
Why is it like that? Why can't it be another way?
Co-founders Doug, Deip Kumar, Anup Chamrajnagar, Anish Chadalavada, and other Gradial employees at the event Show Us Your Craft, in Seattle, August 2026. (Gradial’s LinkedIn)
How do you recognize major innovations around you?
True to his roots, Doug’s opinion is that staying curious, reading, and constantly updating your perspective is what lets you tell, early on, whether something new and disorienting is going to matter or not.
And AI is an accelerant for getting to that frontier.
“You had to walk to the frontier. Now, you have a motorized vehicle … AI can get you to the frontier in almost every subject very quickly, and …. as long as you remember what it's like to learn, it can almost feed you a perfect curriculum of what you need to learn to be effective … Still, (there’s) obviously value in deep expertise, but the ability to catch up to the field has gotten a lot better.”
He wants Gradial to always operate at the frontier, watching where things are headed and building to support that future once it actually arrives.
In other words, to be the fortune teller.
It’s as he did as a toddler, asking why over and over again. But things are different now.
Back then, it meant seeking answers someone was busy answering.
Now, it means writing that answer himself.
1. Don’t try to be everybody’s cup of tea
Trying to create something suitable to serve every industry sounds like the safer bet, but it means serving no one particularly well. Go deep in one domain, understand its workflows well enough to resolve its real pain points, and turn that into a tool that teams can't live without.
2. If they're not banging on your door, ask why (and fix it)
If people aren't chasing after you once you've built something, pay attention. Sometimes the product isn't there yet, and you need to change its course. Other times, nobody knows it exists, and you need to put the word out. The point is, it’s an important signal to pay attention to and to interpret correctly, as soon as possible.
3. Your first edge won't be your last one
Gradial's early advantage was reliable JSON generation, something models couldn't yet produce with consistency, but it didn't last. Once the models caught up, that edge disappeared. Don’t think your current advantage is going to be a permanent one. Be prepared. However, commoditization isn’t the enemy. When your advantage becomes commonplace, that's a signal to move on and find the next thing worth building.
4. Build for what's next, even if you don’t know what it is yet
What more can we offer this year, next year, and the year after? A subscription business demands that value keep compounding, and the only way that happens is if you build with the problems of the future in mind.
5. Get credibility early
A pitch can only get you so far. A critical mass of strong customer references is what makes you take off. Better references lead to better demos, better demos build more expertise in the domain you're going after, and that's what wins the next reference.
