I'm a sales professional who taught himself to build production software with AI.
I run Cardea, where I'm building Enki IQ and Codebrief solo with Claude Code. Now I help other companies do the same compression.
For the last 5 years I sold software for a living. Mostly mid-market B2B SaaS, mostly to revenue and operations teams. I was good at it, and I learned the patterns: which deals close, which never do, what reps do that actually moves the number.
In 2025, I started Cardea and began building Enki IQ. The catch: I'm not a software engineer. I'd never shipped production code. The plan was to find a technical co-founder. The plan didn't survive contact with Claude Code.
I started prototyping with Claude. Then I shipped a beta. Then I built an MCP server, which is currently under review by Anthropic for inclusion in their official marketplace. Then I started selling the product to real customers.
Somewhere along the way, other founders and operators started asking me how I did it. Then they asked if I'd do it for them.
That's the consulting practice. I help companies adopt AI by actually shipping the thing. The compression I did on myself, applied to your team.
Four positions.
Operators beat advisors.
Anyone telling you how to use AI should be using it themselves. Demand proof.
AI-augmented humans beat full automation.
Sales teams that use AI for research, planning, and talking points, while keeping humans in the relationship, will outperform fully automated pipelines. The same logic applies to most knowledge work.
Ship, don't slide.
Every engagement should produce working software, trained people, or both. Not a deck that sits in a Drive folder.
Claude-native by default, model-agnostic in practice.
My deepest expertise is in Claude, Claude Code, and Claude Cowork. I'll work in OpenAI or Gemini stacks when the client needs it, and I'll tell you honestly which one I think is right.