Doug Tallmadge, Co-Founder and CEO of Gradial – Interview Series

Doug Tallmadge, Co-Founder and CEO of Gradial, is a technology entrepreneur and engineering leader whose career spans aerospace, satellite communications, finance, and enterprise AI. Before launching Gradial in 2023, he spent nearly five years at SpaceX, where he helped scale the Starlink network from its earliest stages, led network analysis supporting growth from zero to more than 500,000 customers, and managed teams of software engineers and data scientists working on network optimization and simulation. He also contributed to Starlink’s direct-to-cell satellite initiatives and early satellite power systems. Earlier in his career, Tallmadge held roles at Lockheed Martin and Bridgewater Associates, combining deep technical expertise with experience in large-scale systems and business operations.

Gradial is an enterprise AI company focused on what it calls agentic marketing operations, helping large organizations automate the operational work that sits between creative strategy and campaign execution. Rather than concentrating solely on AI-generated content, Gradial’s platform orchestrates AI agents across systems such as Adobe Experience Manager, Salesforce, Jira, content management platforms, digital asset management tools, and marketing workflows. The platform can author, optimize, tag, quality-check, and deploy marketing assets while enforcing governance, compliance, accessibility, and brand standards. By creating a shared organizational knowledge layer that connects enterprise systems, Gradial aims to reduce the manual coordination that often slows large marketing teams, allowing organizations to launch campaigns faster, scale personalized content production, and manage increasingly complex digital experiences with greater efficiency.

You helped scale Starlink from zero to hundreds of thousands of users at SpaceX, working on complex network systems before founding Gradial. What was the moment when you realized enterprise marketing infrastructure had similar systemic inefficiencies, and how did that insight lead to starting Gradial?

In my previous role, I saw firsthand how friction in large organizations compounds, even the smallest inefficiency can slow down every process that follows. Marketing was where I noticed these bottlenecks had the most negative effect. Teams were spending a lot of time navigating platforms, uploading assets, managing approvals, and coordinating handoffs. Strategic thinking and creative direction were getting buried under layers of execution.

It’s a perfect use case for AI, but what struck me was where the industry’s attention was focused. Most platforms were using AI to write content, but when I talked to marketers at large organizations, lack of content wasn’t the issue. They could produce content easily. What they couldn’t do was publish it quickly. Going from idea to live campaign took days, weeks, sometimes months. Most AI investment was going toward creation, while the harder problem of actually getting content out the door was completely ignored. We built Gradial around that gap: deploying agents that handle everything from drafting and approvals to publishing, brand compliance, and performance optimization, so marketers can focus on work that requires human judgment.

You’ve argued that the traditional “system of record” has effectively become a system of tech debt. What specifically breaks in legacy enterprise marketing stacks when companies try to adopt AI?

A system of record, the CMS, project management tools, approval workflows, and asset libraries was built to organize information, not move fast. When companies bolt AI onto these existing systems, things fall apart quickly. The tool doesn’t know the business it’s serving, the brand voice, compliance requirements, or approval processes. So marketers end up reviewing every output and correcting mistakes caused by missing context. For organizations already managing thousands of pieces of content across multiple channels, bolted-on AI becomes another complicated process to manage. 

Most of the AI conversation in marketing has focused on content generation. Why do you believe the real opportunity lies in execution, and what does that shift look like in practice?

Marketers aren’t struggling to produce ideas. They’re struggling to move content through review cycles, maintain consistent brand voice, work across dozens of tools, and optimize for different audiences.

There’s also a structural shift happening in how brands get discovered. Google searches fell nearly 20% in the past year, which means brands can no longer rely solely on SEO. They need to show up in AI search responses, which requires publishing content that AI models can easily interpret, cite, and surface. The problem is that those models constantly change what they pull from. A brand that appears in an answer today can quietly disappear tomorrow. By the time marketers identify the gap, update the content, and clear approvals, the models have moved on.

Content generation alone doesn’t solve this. Even if you produce content faster, you can still be stuck in the same approval queues, with content that isn’t optimized for AI visibility. Gradial plugs into the tools marketers already use and handles work that usually slows everything down, like content revisions, brand compliance checks, and publishing across platforms. The result is campaigns that ship in hours instead of weeks and content that’s accurate, on-brand, and visible where audiences are actually looking.

Gradial is often described as an agentic platform. Can you walk us through what an AI agent actually does inside a Fortune 500 marketing workflow on a day-to-day basis?

The best answer is what we’ve seen with customers. At T-Mobile, running a campaign meant coordinating across multiple teams, collecting approvals, and waiting on production queues that stretched for weeks. With Gradial, an AI agent handles the entire process from creative brief to reviews and publishing. What used to require a marketer to manually chase down every step just happens. The result was 90% faster time to campaign execution.  

At AWS, every new web page meant briefing an agency, waiting on a draft, going through rounds of revisions, and manually coordinating to get it published. A specialized agent made page creation 20x faster.

One of the biggest challenges in enterprise AI adoption is trust. How does Gradial ensure reliability and governance when AI agents are making execution-level decisions across systems?

Trust is the whole game in enterprise, especially in finance and healthcare, where a compliance mistake has real consequences. Every Gradial deployment starts with giving the agent context on the brand’s voice, messaging guidelines, compliance requirements, approval workflows, and industry standards.

The level of human involvement is fully customizable. Teams new to this or operating in regulated environments can require sign-off on everything before it goes live. And, every action the agent takes is logged with a full audit trail. What we’ve found is that trust builds naturally, with teams that start with tight controls gradually expanding the agent’s autonomy as it proves itself on real work. The goal was never to hand everything over on day one, but to build something that earns more responsibility as it demonstrates it can handle it.

Your platform integrates into existing tools rather than replacing them outright. Is the long-term vision to augment the current stack, or fully replace it?

For now, the focus is on meeting teams where they are. Most enterprise marketing orgs have spent years building workflows around specific tools, like Figma, Adobe, and project management systems, and asking them to abandon that infrastructure isn’t realistic. Gradial plugs in and handles the execution work that’s been slowing those workflows down.

The longer-term vision is more ambitious. A lot of legacy software underneath enterprise marketing was built for a world before AI, and it shows. Systems designed purely to store and organize content will look very different when an agent can hold that same context and act on it simultaneously. We think the distinction between where content lives and where work gets done eventually collapses, and we’re building toward being the infrastructure layer where both happen.

Large enterprises often have deeply entrenched workflows and internal resistance to change. What does it actually take to successfully deploy an AI-native system into a Fortune 500 organization?

The first thing we do is make clear that Gradial isn’t asking anyone to change how they work. That sounds simple, but it’s the biggest hurdle in enterprise AI adoption. Most tools arrive with a new system, a new interface, a new way of doing things, and then wonder why adoption stalls six months later. We start by mapping how a team actually operates and configure Gradial around that.

Adoption sticks when teams see it working on real work quickly. Not a demo, not a sanitized pilot, but their actual campaigns, their actual content, their actual workflows. Once a team sees their publish time go from three weeks to the same day, internal resistance tends to take care of itself.

From a technical perspective, what differentiates an execution engine from a typical generative AI layer?

A generative AI layer gives you a draft, but an execution engine gets it published. Most AI tools for marketing stop at the output, they’ll write the copy or suggest the edit, but the moment that output needs to move through approvals or onto a live page, marketers are left to manage it. That handoff is where most of the time actually goes.

An execution engine is different because it’s embedded in the tools a team already uses and understands how that specific business operates. When a marketer asks the agent to do something, it moves the work through every required step, getting approvals and only surfacing decisions that actually need a human. The marketer stays in control without being in the middle of every step.

Many companies are experimenting with AI, but few are seeing meaningful ROI. Where do you think most organizations are getting it wrong today?

The most common mistake is treating AI as a productivity tool rather than a fix for a broken process. If every output needs heavy editing, or has to be manually copied into another system, or still requires five approvals before anything goes live, you haven’t solved the problem.

The companies seeing real returns got specific. They identified exactly where work was getting stuck and deployed AI against that problem, not broadly, not as an experiment, but with a clear sense of what slow looked like before and what fast should look like after. That precision is what separates real ROI from a tool that impresses in a demo and collects dust six months later.

When you get it right, the results are tangible: customers using Gradial have seen 8x faster campaign activation, 35%+ in operational savings in their first year, and 100% brand and accessibility compliance on every piece of content.

If we fast forward five years, what does a fully AI-native marketing organization look like, and what role does Gradial play in that future?

In five years, the operational layer of marketing will be autonomous. A marketer asks an agent to launch a campaign, and it’s live by the end of the day because the agent handled every step in between. Brand visibility in AI search gets monitored and adjusted automatically. Compliance happens continuously throughout the process rather than as a last-minute check. Marketers spend their time on what they were actually hired to do: building the brand, developing creative, and making strategic calls on customer experiences that require human judgment.

Gradial’s role is getting marketing organizations there first, not by convincing them to rip everything out and start over, but by proving inside their existing workflows that this way of operating is possible. The teams that figure that out early will have a compounding advantage over the ones still running campaigns the old way.

Thank you for the great interview, readers who wish to learn more should visit Gradial.