Most marketing teams keep a list of things they would build if they had an engineer for a week: the tool the sales team keeps asking for, a microsite for one campaign, the report someone assembles by hand every Monday, a translation workflow that stalls at the vendor. I build those. The proof is this site. The Audience Finder searches 950,000 third-party audiences from 287 data providers. The Ad Mockup Generator renders one image into 39 placements. The reference library holds 501 defined terms and 22 platform profiles. The AI, crypto, digital privacy, and PFAS regulation registers track the law by jurisdiction, cited to primary sources. And the homepage, the service pages, and both tools exist in twelve languages, built by a translation pipeline with a style guide and a review pass for every language.
Some of it is visible: calculators, finders, configurators, campaign microsites, and internal utilities, built as plain web software your team can host anywhere. Some of it runs in the background. A script assembles the weekly report from the ad platforms and flags a broken pixel, a paused campaign, or a budget pacing off before a client notices. Content and translation pipelines draft, edit, and localize at volume, with a human editor in the loop and a validator that refuses to publish anything that breaks the rules. And for teams in health, finance, and public affairs that cannot put patient data, deal flow, or client positions into a public model, I set up private and local LLMs.
This is a separate practice from SEO + GEO. That page covers how your brand shows up inside AI answers. This one covers building and running things with AI inside your own operation. The two meet often: the structured data, the machine-readable editions, and the measurement of AI referrals that GEO calls for are usually built with the tooling described here.
The person who plans your media builds the tool. Most internal tool projects die at the specification, when a developer who has never run a campaign builds what the brief says and misses what the team meant. I have twenty years in the work these tools are for, so the first version is usually right about the workflow and the second is the finished one. The scope is marketing and communications: no core product engineering, no systems of record, nothing that belongs with your IT department.
An engagement starts with the workflow you want gone and ends with a working version in your hands, documented so your team owns it. If a tool only makes sense with your data inside a private model, the setup runs on your infrastructure or a dedicated instance, and nothing leaves it.
Name the thing you keep doing by hand
Describe the workflow. The first working version usually follows within days.
Start the conversationQuick answers
What kinds of tools do you build for marketing teams?
Calculators, finders and configurators, campaign microsites, internal utilities, reporting that assembles itself from the ad platforms, QA checks that run on a schedule, and content and translation pipelines with a human editor in the loop. The Audience Finder, the Ad Mockup Generator, and the reference library on this site are the working examples.
Can our data stay private?
Yes. For teams in health, finance, and public affairs the setup runs as a private or local model on your infrastructure or a dedicated instance, so patient data, deal flow, and client positions never reach a public model. Where a public model is fine, I say so and use it.
How is this different from SEO + GEO?
SEO + GEO is about how your brand appears inside the answers ChatGPT, Gemini, and Claude give. AI Tools + Automation is about building and running your own tools and workflows with AI. They share techniques, and the GEO work often depends on tooling built here.
From The Signal
View all notesA Free Reference Library for Digital Advertising, and Five Other Things We Give Away
501 defined terms, 22 platform profiles, sourced benchmarks, current ad specs, calculators and templates. Free, ad-free, no gate, CC BY licensed.
September 18, 2026One Image In, 39 Placements Out. A Free Ad Mockup Tool Built to Current Specs.
One square and four lines of copy become 39 ad mockups across every major channel, rendered to current specs and exported for a deck. Free, in the browser.
July 17, 2026How This Site Is Built to Be Read by Machines and Humans
A colophon for the AI search era: hand-built static HTML, consent-gated measurement in Europe, machine-readable editions, thirty-plus checks before any deploy.
July 15, 2026The Right AI Model Is the Cheapest One That Clears the Bar
Claude, ChatGPT, Gemini, or Perplexity. Fast or reasoning model. Choosing an AI model is a media buying question, and most people answer it by habit.
July 7, 2026Your AI Subscription Is Worth $25,000 a Month. You're Using It Like It's Worth $20.
Claude, ChatGPT, Perplexity, and Gemini all moved to compute-metered usage windows within a year. The playbook for extracting what you already pay for.
June 9, 2026The Media Review Meeting Is Where Insight Goes to Die
Forty slides, every number green, no decisions made. The monthly media review is broken the same way at most companies, and fixing it costs nothing but nerve.
