The Investment Thesis
Build up CPG recurring revenue streams for physical product — in the consumer packaged goods space, a running supply chain typically values at 0.8x revenue, a fast-growing one at up to 2x. Then layer proprietary artificial intelligence software on top, which commands a 5–9x multiplier. The combination would result in a company valuation of approximately $40 million.
That was the thesis. As 49% owner and co-equal manager — one of two on the board of directors — I set out to build both sides of that equation simultaneously: the physical product business and the AI platform that would make it defensible.
The Build: Concept to Cash
The first two years were about learning CPG from the inside. I embedded myself in the entire development workflow from concept to cash:
Reverse-engineering top sellers. Take a top-selling product — usually a grocery store house brand — and understand exactly why it sells.
Clean-label reformulation. Replace all the dyes, additives, and harmful ingredients with clean-label alternatives: better nutrition, healthier ingredients, recyclable materials. The stuff that gives people cancer and affects kids’ development gets removed. The product gets better, not just different.
Supply chain optimization. Optimize the entire chain from ingredient sourcing through manufacturing through distribution. Every link in the chain is an opportunity for margin improvement or margin leakage.
I sat side by side through this entire process. As an AI engineer and business leader simultaneously, whenever I saw an inefficiency — a manual process, a risk someone was introducing, a bottleneck that didn’t need to exist — I would write an application for it. This is where the Lean Six Sigma methodology from Oracle and Purdue came in: I applied the same statistical process control discipline that reduced Oracle’s outage minutes by 58% to my own business operations.
National Retail at Scale
The results of that build process:
- $9M ARR in just over three years, from zero
- 10 SKUs on Sprouts shelves across 24 states, plus other national and regional retailers
- Zero products with Red 40, Yellow 5, or other harmful dyes
- 100% ingredient traceability — we know every farm
Getting a product onto a national retailer’s shelf is extraordinarily hard. Getting 10 SKUs across Sprouts stores in 24 states in three years — while maintaining clean-label standards and FDA compliance — required serious operational discipline.
My co-founder established the manufacturing partnerships and built out the production network. I came in and solidified those relationships through quality engineering — applying the same CAPA and statistical process control methodologies I’d honed at Oracle to ensure consistency, compliance, and reliability across every co-manufacturer and supplier in the chain.
The ERP Survival Story
In the last year, I inherited the full weight of supply chain operations: material moving across America, distribution, payments, controller functions. I would wake up at 4:30 AM in a cold sweat convinced I was going to lose my house that day.
Fraud was endemic in the business — inherent to the CPG supply chain model with its web of co-manufacturers, ingredient suppliers, brokers, and 3PL partners. Existing systems couldn’t catch it. Existing processes couldn’t prevent it.
So I wrote my own ERP system. Always Cool AI wasn’t a nice-to-have automation project — it was a survival tool. Built from the ground up to remove risk from supply chain operations, provide real-time oversight across every transaction, and catch the fraud that was inherent in the business. Automated purchase order processing, inventory management, and broker communications.
The AI Supply Chain Optimizer project page covers the technical platform in detail — 8 specialized agents, LangGraph orchestration, FDA compliance automation. This page is the business story behind why it got built.
The CAPA Bridge: Oracle to FDA
The most unexpected connection in my career: the CAPA program (Corrective Action Preventive Action) I learned in depth at Oracle turned out to be FDA methodology. I’d inherited one of the most degraded platforms in the world — Gen1 and Gen2 — and turned it around using CAPA. When I left Oracle to co-found ACB, I brought that methodology with me and discovered it was the exact same framework the FDA-regulated food industry uses.
I’d learned FDA methodology at Oracle without knowing it was FDA methodology. The regulated industries thread isn’t something I constructed after the fact — it’s something that’s been running through my career from ID Analytics through Oracle through ACB and into my current practice.
The pipeline: ID Analytics statistical analysis → Purdue Statistical Process Control → Oracle CAPA → FDA compliance at ACB → NRC/FINRA governance frameworks. The math started at ID Analytics.
Engineering, Finance, and Full P&L
My role at ACB wasn’t just technology. I led engineering and finance. Was on every product design call from concept through manufacturing. Managed the controller function and the accounting relationships. Solidified manufacturing partnerships, cold chain logistics, and broker/distributor networks through quality engineering discipline.
Running full P&L for a physical product company is fundamentally different from managing technology budgets. At Oracle, I had a $70 million annual budget managing hardware, firmware, and operations together — when something broke, you were coordinating across physical equipment, vendor firmware updates, and operational teams simultaneously. At ACB, the pressure was different but equally visceral: pallets of product sitting in a warehouse accruing storage charges while you figured it out. The cash flow pressure of physical goods — where inventory ties up capital and every day of delay costs real money — taught me lessons that even large-scale infrastructure roles never could.
Lessons Learned
1. Lean Six Sigma transfers across industries. The same SPC discipline that reduced Oracle outages by 58% applies to CPG supply chain variance. Process control is process control.
2. AI in regulated industries starts with compliance. It’s easier to make a compliant system fast than a fast system compliant. FDA doesn’t care about your velocity.
3. Fraud detection is a survival skill. The supply chain fraud patterns I encountered at ACB directly informed FrawdBot and the financial intelligence bridge that connects behavioral detection to forensic accounting.
4. The CPG + AI valuation thesis is real. Recurring physical product revenue provides a stable base. Proprietary AI creates the multiplier. The combination is more defensible than either alone.
5. Co-founding is not consulting. When it’s your money, your name, and your sleepless nights, every decision has weight. Three years of building ACB taught me more about operational leadership than a decade of enterprise employment.