I help businesses move from BI dashboards to AI that does the work.
Seven years as a founder and operator in CPG analytics. Now building AI-native ventures and a personal AI operating system. Based in Denver.
AI automation isn’t a tool you bolt on. It’s a system you build.
Most companies trying to use AI today are wiring up one-off integrations — ChatGPT inside a Zapier flow, an AI feature in a SaaS tool, a chatbot on the website. These work for a week, then break the moment a process changes or a customer asks a question the script didn’t anticipate. The result is a graveyard of half-finished pilots and a leadership team that’s lost faith in “AI” as a category.
The companies winning right now are doing something different. They’re treating AI as a new layer of their operating system — with the same care they’d give any production system. Pieces that compose. Approvals that scale down as trust earns up. Knowledge that outlives any vendor. Hard rules where judgment really matters.
I’ve built this twice. As Head of Product at Daasity, I led the transformation of the company’s analytics platform from a traditional BI tool into a multi-agent AI system serving Fortune-500 CPG brands — a production deployment with zero hallucination tolerance, a semantic layer at its core, and 200+ pre-approved analytical patterns. After leaving Daasity in late 2025, I rebuilt the same patterns from scratch at personal and small-business scale, to test what actually works without the safety net of a 50-person engineering org.
The projects below are the answer. A personal AI operating system that runs my work and my knowledge. A six-agent operations stack running a real Shopify business. A content pipeline turning 876 hours of dharma teachings into publishable books. A portable analytics package that turns a raw retail data file into a competitive analysis deck — installable in five minutes.
Every project ships with a downloadable package. Clone, install, run.
Projects
chris-os
A personal AI operating system that runs my work, knowledge, and decisions across desktop and cloud.
tibetan-spirit-ops
Six AI agents running a real Shopify business — fulfillment, inventory, customer service, marketing, catalog, finance.
cpg-agents-2.0
Drop in a raw Nielsen, SPINS, or Circana export. Get a competitive analysis deck. Install in five minutes.
dharma-writer
A 4-skill content pipeline turning 876 hours of dharma teachings into publishable articles and book manuscripts.
ace-ai
AI-native construction project management. Upload a spec PDF, get a project plan with schedule, risk assessment, and client-ready deliverables.
bloomy
A generative ambient music app for iOS inspired by Brian Eno's Bloom. Touch to create evolving soundscapes, or let it compose autonomously.
Four very different projects. One pattern underneath all of them.
Modularity that compounds
Every project breaks down into small, portable pieces — the way a cookbook is composed of recipes, not one giant document. A piece written for one project can drop into another tomorrow with no rewrite. The practical result is that my second project moved twice as fast as my first, and my fourth moved twice as fast as my second.
Knowledge that outlives any vendor
Notes, decisions, customer interactions, training material — all stored as plain text files in folders. Databases are used only to make them searchable, never as the source of truth. If every piece of software I use disappeared tomorrow, my knowledge is still readable in any text editor.
AI earns trust the way a new employee does
Every workflow starts requiring human approval on every output. The system tracks how often a human edits or rejects the AI’s work. Once the override rate drops below a threshold and stays there for 30 days, the workflow earns the right to act faster. This is the missing pattern in most enterprise AI rollouts.
Hard rules where AI doesn’t get to decide
Some decisions are non-negotiable: don’t send a customer a spiritual teaching, don’t auto-execute a financial trade, don’t translate sacred terms. Those are coded as deterministic rules before any AI reasoning runs. The model handles the long tail of judgment; the rules handle the cliffs.
Background
Seven years as a founder and operator in CPG analytics. Started my career in category management at brands including Justin’s and I and Love and You. Founded Red Fox Analytics in 2017 — a services business delivering workflow automation, analytics, BI, and data management for CPG brands. Ran it for five years, scaling a small team and serving dozens of consumer brands.
In 2022, Red Fox was acquired by Daasity, the leading analytics platform for D2C consumer brands. I joined as Head of Product and spent the next two years leading the platform’s transformation from a traditional BI tool into a multi-agent AI system — production deployment, zero hallucination tolerance, semantic layer at the core. Departed in November 2025 to build full-time.
Davidson College ’15, Religious Studies (Buddhism). Currently building ACE AI (AI-native construction PM), Tibetan Spirit (Shopify D2C), and a personal AI operating system. Seeking roles at AI-native companies or operator/advisor seats helping enterprises enter the agentic era.