What does an engagement cost?
Company Brain starts at $3,500 setup plus a retainer starting at $500/month. Workflow sprints start at $5,000 fixed. Systems retainers start at $500/month after a first system ships. Every public number is a starting-at floor for standard scope. You'll get a fixed quote in writing before you commit.
Why is Company Brain “starting at” $3,500?
That price is the standard build: core sources (typically Slack, Google Drive, and meeting transcripts), one workspace, cited answers, permissions, deploy, and handoff. More data sources, one-off or custom ingestion, heavy history, multi-client permission models, or extra surfaces are inventoried on the call and quoted as a higher fixed price. You're never billed hourly by surprise.
Who owns the system when we're done?
You do. Code, accounts, data, and documentation live in infrastructure you control. Every engagement ends with operator docs, a training walkthrough, and QA checklists so the system runs without me. The retainer is optional, not a hostage situation.
How is this different from hiring an AI agency?
Two ways. First, scope: you buy a defined deliverable at a fixed price, not an open-ended retainer. Second, depth: I spent fifteen years running growth channels before building systems for them, so the tools fit how marketing teams actually work because I've done the work they automate.
Will you tell me if AI won't help?
Yes, on the first call, for free. Some workflows aren't ready: the data is too messy, the process too inconsistent, or the cost of an error too high. Building those anyway is how AI projects fail. If yours is one of them, you'll hear it from me before you spend anything.
Do you still do paid media?
By request, for clients who already trust the systems work or come in warm. Primary work is AI systems: knowledge bots, sprints, reporting agents, and internal tools. Growth and measurement leadership is available when it fits; it is not the main offer on this site.
What do you build with? Which AI models?
Model-agnostic, on purpose. Each system uses the right model for its job: Anthropic's Claude for one workload, OpenAI for another, open-source models running privately when data can't leave your infrastructure. Around the models it's production-grade, boring-on-purpose tooling: Python and the systems you already run, like Slack, Google Workspace, Shopify, your ad platforms, and your CRM. When a better model ships, your system can switch to it. You're never locked to one vendor's roadmap.
Do you replace our team or our agency?
No. The systems remove the manual, repetitive layer so your people spend their time on judgment calls. Agencies hire me to build their internal tooling for exactly this reason: it makes their senior people more valuable, not redundant.
