GTM engineer · senior media buyer
A GTM engineer who also runs the ads.
I'm Michael O'Brien. I've run marketing for 300+ brands over 13+ years and managed $50M+ in ad spend. These days the job is as much engineering as buying: I build my own trackers, media-buying tools, MCP servers and API connections, and run them with AI agents in Claude Code and Codex.
Free. Bring the account, and your stack, and I'll show you where it's leaking.
What GTM engineering means here
It's the plumbing between the ad and the sale: tracking, data, lead routing, landing pages and the tools that tie them together. Most teams split it between a media buyer, a developer and an analyst. I do it myself and buy the media too, so the tools get built around what actually moves revenue.
- TrackingServer-side GTM, Conversions API and the data layer, with one event per lead and each CRM stage sent back to Meta and Google.
- DataAd spend joined to tracked revenue in BigQuery and Looker Studio, read by first click, last click and modeled attribution.
- Lead opsCustom and partner lead APIs, routing and enrichment on n8n, Make and Apify, into GoHighLevel or HubSpot.
- PagesLanders I code myself, with A/B tests, shipped on Netlify or Vercel with git.
- ToolsMy own media-buying tools, MCP servers and API connections, including report pullers for Meta, Google and Microsoft Ads.
Run from the terminal, with AI agents.
Plugs into what you already have
- Company MCP serverHave one? I connect to it and work straight from the terminal.
- Ad accountsMeta, Google Ads and Microsoft Ads over MCP, TikTok by API
- Tracking and analyticsPostHog, GA4, the GTM data layer, server-side GTM, Hyros, Triple Whale
- CRM and leadsGoHighLevel, HubSpot, custom and partner lead APIs
- DataBigQuery, Looker Studio and the pipelines you already run
- Automationn8n, Make, Apify
Already have pipelines? I use them as they are. If something's missing, I build it.
What the agents do
- Spend against revenue, every day. Meta, Google and Microsoft spend joined to tracked revenue, with anything past breakeven flagged.
- Tracking checks. Pixel, Conversions API and CRM events compared, so duplicates and silent failures get caught early.
- Creative at volume. Competitor ads turned into hooks, images, video, copy and a test plan, every batch checked against your claim rules.
- Page research and builds. 10 to 20+ live pages studied before I build yours, then shipped with git.
- Lead ops. Leads enriched and routed, and each CRM stage sent back to Meta and Google.
The agents do the legwork. The calls on budget, creative and offer stay mine.
13+ years$50M+ in ad spend managed300+ brandstwo exits. See the results and screenshots.
Straight answers
Is the AI a separate service?
No. It’s how I run the work. You hire me for growth: tracking, creative, pages and buying. The agents and the tools I’ve built are how one person does all of it.
Can you work inside our company MCP, CRM or data warehouse?
Yes. If you have a company MCP server, I connect to it and work straight from the terminal. The same goes for your ad accounts, PostHog, GoHighLevel, BigQuery and the pipelines you already run: I use them as they are. If something’s missing, I can usually build it.
Who actually does the work?
Me. The agents do the legwork: pulling data, checking tracking, researching pages and making creative at volume. The calls on budget, creative and offer stay mine, and I’m the one on your calls.
Start with the free full-stack audit.
I look at the ads, the page, the tracking and the data together, and tell you where it's breaking.
Need marketing now? Hire me to start.