I market technical products, and I run that marketing on AI agents I built and maintain.

The Pitch: How I'd Start

GetBlock launched Crypto AML, the Platform API and Market SQL in about a month, and each one has to reach the few people who pick an RPC provider for their engineers. I would start by writing down what those people ask before they choose, such as which provider is fastest for their chain and region, and how limits and pricing behave at their traffic. Each week, I would track which providers Google and AI answers cite when buyers ask those questions. GetBlock already ranks well in organic search and AI answers.

For each launch I would agree one positioning statement with the product team, then write the X thread, blog article and partner announcement from it. I would turn that process into a team skill, then edit every draft by hand before it ships. Campaigns would be judged on signups from accounts that build on GetBlock, and spend would move toward whatever brings those in.

The Proof: Workflows I Built and Run

I built 81 agent skills for my own marketing work and published 34 of them as Marketing Agent Skills on GitHub.

I built an AI system in Claude Code that runs paid acquisition and measurement for my own business and our clients. It reads the funnel from creative to checkout and reconciles what each ad platform reports against real sales. It caught the Meta pixel crediting campaigns with sales from other products, and on one cold campaign it showed that buyers dropped at checkout, so the fix was the checkout and not the ads.

I also wrote an AI answer visibility audit in Python that asks AI search the questions engineers ask and records who gets cited, with the code public. And I built the Prospect & Outbound Engine, 40+ Python modules that took one live run from 409 sourced agencies to 267 send-ready leads.

The Proof: Technical Buyers and Owned Media

At Reform, I repositioned the product around lead conversion for a technical and business audience, and trial activation rose 34%. The Feature Launch Plan Framework I wrote there lifted new feature adoption 26% within 90 days. I also built the partner program on PartnerStack, with co-marketing playbooks alongside Clay, Lovable and Claude Code.

I write everything I publish. The Marketing In Action newsletter goes out weekly to 5,000+ subscribers across 86 countries at a 47.5% average open rate.

For developer and infrastructure buyers, I wrote a draft technical case study for Cosine, an AI coding agent, and rewrote a comparison page for Dash0, an observability platform. In crypto, I published a search and AI answer teardown of Binance across six African markets in four languages.

Let's talk about GetBlock.

If this is the right problem for you, I'd like to talk about it directly.