Director of Marketing Who Embedded AI Into 80% Of Operations Says You Must Treat AI Agents Like Staff
Learn how a marketing leader uses AI agents like staff to scale operations, save time, and improve output, with real workflows and lessons on where humans still lead.
AI Integration: George Giles has successfully embedded AI into over 80% of Foretellix's marketing operations.
AI as Staff: Giles recommends treating AI agents and prompts as company staff with defined tasks and responsibilities.
SEO and AEO: AI improves SEO by developing content strategies based on unique prompts, enhancing search visibility.
Human Touch: Despite AI's capabilities, humans must still handle branding and messaging for nuanced understanding.
Leadership Support: Early leadership buy-in is crucial for successful AI adoption and integration in marketing.
George Giles' career has been focused on scaling tech companies. He has worked with HP, Intel Ignite, DocuSign, and many other well-known brands, and he's currently the Director of Marketing at Foretellix, where he has embedded AI into more than 80% of marketing operations.
We sat down with him to get a sense of what's possible with AI in marketing roles. Here's what he had to say.
The Mind Of A Developer And Soul Of A Marketer
I'm a technical visionary bridging the gap between deep-tech engineering and high-performance marketing. I've spent my career launching and scaling technology companies across the U.S. and global markets. I have worked with global titans and innovators such as HP, Intel Ignite, HPE, DocuSign, Trend Micro, and Lumenis. And I have two successful exits under my belt as a founder.
Want more from The CMO?
Sign up for a free membership to complete reading this article:
Have an account? Log In
I'm currently the Director of Marketing at Foretellix, a leader in the autonomous vehicle industry. Beyond the boardroom, I'm an AI advocate and mentor. I volunteer my time teaching practical AI applications to help marketing teams bridge the adoption gap, focusing on building custom agents and tools that move the needle on ROI rather than just adding to the noise.
My dual DNA sets me apart — I have the mind of a developer and the soul of a marketer. Because of that, I'm an early adopter. I've been building with AI for years. In fact, I've embedded AI into over 80% of all marketing projects and operations at my organization, moving beyond experimentation to full-scale operational excellence. That's a rare feat in the industry.
Setting The Stage
At Foretellix, our marketing defines the standard for safety in the autonomous vehicle revolution. We operate in the Physical AI space, providing the verification and validation (V&V) toolchain global automakers use to prove their self-driving systems are safe for public roads.
Here's the context:
- Audience: We market to an extremely sophisticated, high-barrier-to-entry audience — a complex committee of C-suite executives at global OEMs and Tier-1 suppliers.
- Reach: We manage a multi-regional strategy across the U.S., Europe, and Asia, ensuring our messaging aligns across multiple markets and languages.
- Ecosystem: Co-marketing with industry titans is a core part of our strategy. We don't just exist in a silo; we integrate and go to market alongside partners like NVIDIA.
- Channel strategy: Technical authority at scale. Because our product solves a multi-billion-dollar engineering hurdle, our marketing channels must prioritize Technical Authority (no B.S.). That includes:
- Strategic content: We move beyond traditional lead-gen, producing high-level industry content that educates and informs engineers about the latest technology capabilities.
- Experiential high-touch: We leverage major global trade shows, like CES, not just for branding, but for live demonstrations proving our value proposition to the world's most skeptical engineers.
This organization is unique because we "eat our own dog food." Just as our product uses AI to automate vehicle testing, our marketing operations fully embrace AI. This isn't just about using LLMs for copy; it's about using AI agents as an extension of our marketing team. We manage AI agents like staff.
Why Leaders Must Think Of AI Agents As Staff
The biggest mistake most marketing leaders make is treating AI like a better version of Google Search or a faster copywriter.
Instead, start thinking about AI agents and prompts as staff. I like to think of this as a shift from Software-as-a-Service (SaaS) to "Staff-as-a-Service."
The biggest mistake most marketing leaders make is treating AI like a better version of Google Search or a faster copywriter. Start thinking about AI agents and prompts as staff.
Rather than simply using AI, we manage it with clear definitions of its tasks and responsibilities. We bake in feedback, testing, growth, and expansion of responsibilities.
Because agents require time and training to get to a point where they can be trusted to create good results, and even more time for great results. Once you adopt the mindset that they are like staff, you tend to invest more resources in providing quality data to train them, offering meaningful feedback, and finding new improvements continually.
These agents become hugely valuable, saving time and resources.
How Real-World AI Workflows Save Time
To date, I've onboarded 217 specialized AI "staff members" to handle distinct, high-repetition tasks. I treat each agent as a dedicated team member with a specific job description, ranging from messaging research, market gap analysis, content creation, image creation, video creation, to hyper-localized content adaptation for global markets in localized languages.
Here's a simple example of one of these agents. Our "Boss Agent" saves significant time with content creation. I created an agent as a reviewer modeled after a manager's background, experiences, roles, and comments from other content that we worked on. This allowed the agent to review a draft in a very similar way that the manager would, giving us feedback and notes to improve it before it even reached him. Now, he spends less time in review.
How AI Can Be Used to Enhance SEO and AEO Strategies
And there are more complex use cases, of course.
The majority of earned clicks for companies still come from search. The advent of no-click and AI answers means both SEO and AEO are important.
I wanted us to appear in the answers people searched for, but many years of SEO work taught me that I needed to choose the right questions, prompts, and search inputs. Keyword reports weren't a good starting point for choosing prompts, so I started a project interviewing carefully-selected people inside and outside the company to learn how they approached and used AI, and what questions they asked in their research.
Was it a technical issue they were looking for? A known brand? A technology? We needed that answer to determine the prompts. We grouped each interviewee into a category, then created an agent for the interview-script writing process. This agent incorporated the backgrounds of three major groups (technical, sales, and customer-facing), and then wrote the scripts to address their backgrounds and usage. We used Gemini with ChatGPT for the sanity check.
The answers were always surprising. Every time, people told me they weren't the best person to talk about AI, but then they showed me what they were doing with it, and I always learned at least one gem. It turns out prompts are even more unique than search terms, and they're more open.
We then analyzed the interviews and created a list of prompts. Then, we used those prompts to create content strategy for AEO. From there, AmIonAI.com monitored — and still monitors — our weekly progress.
How To Design Effective AI Podcast Workflows
Another workflow was creating an overview podcast with audio and video created automatically from blog posts. My initial workflow was taking content from the blog -> NotebookLM for the audio -> Riverside pulling the transcript -> CapCut for video, audio, EQ, and transcripts.
It didn't work. NotebookLM is great, but the flow had too many steps.
Ultimately, we rethought the steps and dramatically reduced them with a new, effective flow starting with blog -> NotebookLM for the audio -> Descript.
Descript performed all the functions we wanted Riverside and CapCut to do, including text editing. This flow has saved us a ton of time.
Why Branding And Messaging Must Remain Human
So, we use AI for a lot of tasks, but not everything should be offloaded.
Human marketers must still firmly manage messaging and branding. Full stop.
AI is a great tool for brainstorming, but it isn't quite there when it comes to understanding nuance.
There are other places where AI underperforms as well. Most marketing teams spend a lot of time on presentations, for example. I would love for AI to take my corporate template and build new decks from events, demos, customers, etc., with all the graphics, while maintaining our brand book.
So far, though, the tools I've tried — Claude and Canva — aren't there yet. They can't build presentations based on my templates or even outline the talking points on each slide.
Why Gemini Is A Must-Have Marketing Tool
Gemini is my go-to marketing tool.
Writing, brainstorming, research, analysis, strategy work, automated image creation for social media, image editing and more. In the last two updates, Gemini has started creating better answers without a back-and-forth.
It still has factual hallucinations, but it outperforms ChatGPT for my work.
Of course, this could change next week! Last year, ChatGPT was my tool of choice. It's important to keep trying tools and shifting work tasks onto new tools to get the best results.
How Leadership Buy-in Drives AI Adoption
Change management is key to AI adoption. So is leadership buy-in.
Leadership at our organization adopted AI very early, and thanks to that vision, bringing AI into marketing was supported and championed. Without leadership's early buy-in, we wouldn't be nearly so far ahead.
Leadership at our organization adopted AI very early, and thanks to that vision, bringing AI into marketing was supported and championed. Without leadership’s early buy-in we wouldn’t be nearly so far ahead.
I think it's important to know that a few bad experiences can easily taint the future use of AI in an organization. It's important for leaders — and marketing leaders in particular — to keep learning. And remember that AI systems are getting better every day. The speed of change with AI is such that a year is like a lifetime for other technologies.
Why AI Can't Replace Marketing Pros
Some believe AI can replace a marketing expert. I believe it can't.
Effective marketing combines many skills and talents. It requires creativity, humor, analytical thinking, psychology, technical skills, and more. It also requires wisdom from experience.
AI doesn't learn from real experiences.
It has never launched a $1 million Super Bowl ad and failed. That experience never shaped its worldview. It never had to get marketing jobs to pay the mortgage or work on failed products that didn't gain market traction.
These experiences shape marketers, but they haven't shaped AI yet.
Curiosity Drives Marketing Innovation
My advice? Be curious!
Play with these tools to see what they can do. Talk to your team to get more ideas. Does this save us time? Does this elevate our work? Are we moving faster with this? Start with curiosity. Then adopt, test, and refine.
Follow along
You can follow George Giles' work on LinkedIn as he continues testing what's possible in marketing automation.
More expert interviews to come on The CMO Club!
