Wiley Jones

Wiley Jones is the CEO and co-founder of Doss, an AI-native operations platform for companies that move physical products. It covers procurement, inventory and order management on a data model each company can adapt. Wiley's background is in hardware engineering. He left his job when he saw GPT-2 become GPT-3, and Doss incorporated a month before ChatGPT launched, about four years before he joined us in Episode 44 (May 29, 2026).

What Doss does

Making hardware, Wiley kept running into enterprise software that forced the business to operate in ways it shouldn't have to. His one-liner (02:13) is that Doss lets companies adapt their applications to their strategy instead of bending their strategy into the shape of the applications.

He avoids calling it an ERP. Doss calls it an ARP, an adaptive resource platform: core applications for procurement, inventory and order management with simple building blocks underneath. In short, he said, Doss helps customers manage the flow of goods, dollars and data.

The demo

From 03:00, Wiley walked through a demo instance for a fake coffee company.

• A custom SKU model (04:19). The company tracks grinds, including 25 variants of medium grind coffee. Doss handles that with a dimensional data model, and new SKUs come in through a form that feels like Google Forms but writes to the system's database.
• Search across a million orders (05:48). He typed his name into a table of a million sales orders and got matches in a couple hundred milliseconds.
• A built-in agent (06:04). He asked for the last 10 wholesale orders and it pulled them. He compared the agent to Claude Code or Cursor: it fans out across workflows, the data model and sampled tables, and describes the system the way an agent describes a code base.
• The harder question (15:44). He also asked how SKUs, the product catalog, sales channels, integrations and fulfillment connect. When it finished, it described the order process line by line, from sales order to leaving the warehouse, down to buying shipping labels through ShipStation and printing dummy labels for local delivery.

So very, very detailed precise understanding of the system behavior and it's not reading documentation. It's actually reading its own implementation.

— Wiley Jones, Episode 44

Because the data model lives at the application layer rather than in code, Wiley said the agent gets better every time the underlying models do, without more prompt engineering. Jordan called the result exposed tribal knowledge; Wiley's framing was implicit system behavior made explicit.

Key ideas from the conversation

• Narrow the ICP (09:45). Investors like hearing that the market is every company, but Wiley said you need to be very specific. The target is physical-product companies doing $20 million to a couple hundred million in revenue that are focused on winning their market.
• Maturity matters more than industry (10:35). Smaller companies think in binary terms: grow revenue, hit targets. Bigger ones are "enterprise shaped" and act more like big companies in other industries than smaller ones in their own.
• Switching costs mean bad qualification (13:04). Roughly 5 to 10% of the market wants to move each year. The job is to be in those conversations, and in front of companies going through growth or material change.

Number one is that if you're fighting switching costs, then you are not qualifying customers properly.

— Wiley Jones, Episode 44

In the news segment (18:18), on Robinhood letting AI agents trade stocks, Jordan doubted it adds alpha if everyone has the same AI and pointed back to our Kalshi prediction market trading bot and our episode with Public. Wiley thought a natural-language interface for investing could be useful but worried about a more frictionless path to destructive trades.

For more on AI in operations, see AI in logistics, AI in retail and ecommerce and AI in manufacturing and hardware.

FAQ

Who is Wiley Jones?

Wiley Jones is CEO and co-founder of Doss. His background is in hardware engineering; he left his job after seeing the jump from GPT-2 to GPT-3 and incorporated Doss a month before ChatGPT launched, about four years before his May 2026 appearance.

What is Doss?

Doss is an AI-native operations platform that Wiley calls an adaptive resource platform rather than an ERP. It covers procurement, inventory management and order management for companies moving physical products, on a data model customers can adapt.

Who is Doss's ideal customer?

Wiley said Doss targets physical-product operating companies doing roughly $20 million to a couple hundred million in revenue that are focused on winning in their market.

How do you overcome switching costs in enterprise software?

Wiley's answer is that if you're fighting switching costs, you're not qualifying customers properly. He focuses on the roughly 5 to 10 percent of the market that wants to move each year and on businesses going through material change.