The newest hire at Canadian furniture brand Sundays is named Wilson. He’s a logistics powerhouse, capable of juggling freight quoting, booking, tracking and carrier coordination, while also managing damage claims and delays. He’s available 24/7 via phone, email, Teams or Slack. He doesn’t take up any desk space—and if you haven’t guessed it yet, he’s an AI assistant.
Built by San Francisco–based startup Cartage, Wilson is the brainchild of CEO Abdul Basharat and CTO Josh Lampen, who established the company together in 2024. Designed to be integrated within a company’s existing logistics workflow, the AI agent was created to autonomously handle all of the communications necessary for shipping and logistics—sending emails and taking phone calls just as any employee would.
“Customers can call Wilson—he picks up immediately,” says Basharat. “He can answer your question about your shipment status or your delivery updates, and then he can forward it to a person if it’s a more complex question.”
The co-founders came up with the idea for Wilson while working together at logistics software company Rose Rocket. There, they discovered a key problem: Every shipping provider operated with a different software or portal to communicate with customers—and customers, unable or unwilling to track down the proper point of information for each provider, often simply reverted to phone call or email, leading to a messy trail of correspondence.
“That’s why the supply chain is so Wild West,” says Basharat. “When AI models came out, we were like, ‘This entire thing can get automated without changing anything, because AI can now do the emails, the phone calls, turn unstructured data into structured data.’”
In service of that mission, Cartage has managed to drum up significant support among investors. Since completing Y Combinator (the same startup accelerator that got Airbnb and Reddit off the ground), the company has raised close to $10 million, and is currently working on its Series A. Meanwhile, it has also built up a stable of brands utilizing Wilson—including, in the home category, Sundays and mattress brand Saatva. (Those clients pay 25 to 35 percent of what they previously spent on logistics coordination for the company’s services, according to Basharat.)
From the beginning, the company’s goal was to make freight coordination autonomous, which kicked off as a pilot program with customers that resulted in the debut of Wilson roughly a year later. Though the early trial began with a cohort of lumber companies, it quickly became clear to Basharat and the Cartage team that consumer-facing brands—especially in the home furnishings sector—were a better fit for their service. Today, furniture is the company’s biggest market, primarily due to the category’s unique shipping challenges: Unlike lumber or uniformly packaged consumer goods, home furnishings vary in size and shape, making them trickier to transport.
When it came to developing Wilson, Basharat had another goal that is integral to his company’s success: Make AI less scary for people at every stage of the customer journey.
When designing Wilson, the Cartage team sought to make the tool as similar to a person as possible, which is why it comes equipped with its own phone number, email address and remote desktop, all accessible by a brand’s employees. “I know it’s not a person, but that’s the easiest way to understand it,” says Basharat. “It’s actually not that different from hiring: You give them an email, you have to teach them how you do things today. There’s an onboarding process, there’s a learning process as they get up to speed. We found in our sales process that once we really simplified it in this manner, it was easy for people to understand and then start buying and implementing.”
That onboarding process consists of Cartage consulting with a brand’s logistics team and creating a workflow based off of their processes, then coming up with recommendations for how Wilson can help. After the brand signs off on the onboarding document, Cartage creates a number of simulations and scenarios that train Wilson to address the brand’s unique needs, before getting the tool officially up and running.
Once Wilson gets to work, the Cartage team continues to monitor everything the tool does—along with employing an “internal Wilson, watching all of the Wilsons.” All in all, the agent is largely self-sufficient: Basharat says that 88 percent of shipments are handled end to end by Wilson without any intervention; about 10 percent require intervention on the brand side (though only 0.3 percent of those are actual mistakes—most are typically data corrections or customer updates); and 2.2 percent require the Cartage team to step in.
That level of accuracy can be attributed to Cartage’s underlying AI model. Rather than using an “AI wrapper”—a common practice in which an AI startup uses underlying technology from larger models like Anthropic or OpenAI—the company has developed its own model, called Cartage_1, which has been trained specifically on logistics work.
“ChatGPT is basically built for consumers; Anthropic is building more of a coding agent; but nobody’s building a logistics agent,” says Basharat. “It wasn’t until we had enough data—because we had been around long enough that we could use that data—to actually train our own model that we started to make more breakthroughs.”
The company has trained Wilson on thousands of data sets involving real, anonymized logistics scenarios, allowing the agent to handle a significant volume of work on its own. Despite the AI agent taking over the majority of interactions with logistics partners and customers, brands that have employed Wilson say the new hire hasn’t generated much of a reaction among that cohort.
“There hasn’t been [an external response,] and the reason for that is Wilson integrates into our ecosystems like an agent we just onboarded,” says Carl Riano, vice president of operations at Saatva. “Whether it’s a fulfillment analyst, fulfillment agent, retail leader, anyone, Wilson can provide that information. There isn’t a noticeable, ‘Was this AI or was this a human?’”
When Sundays first encountered Cartage in the spring of 2025, the furniture brand was struggling with logistics. Up to that point, it typically chose carriers based on region or type of business, defaulting to those it had previously worked with, rather than embarking on the time-intensive process of searching for the best price or service for a given order. While the company’s leadership knew that it would be optimal to quote various carriers for each individual shipment, it was ultimately too time-consuming. Even a transportation management system, or TMS, often demands a great deal of manual work and spreadsheet-keeping that the Sundays team didn’t have the bandwidth to maintain.
Another sticking point involved dealing with multiple carriers within the supply chain. A piece may start in a warehouse, before getting transferred to a middle-mile carrier, which transfers it to a final-mile carrier, which then books white-glove delivery with the customer. For Sundays, it was hard to know where a product was located at any given time throughout the process, which made it difficult to give customers tracking information or advance planning for delivery. In those conditions, the company wasn’t in the practice of tracking each order’s whereabouts—it only did so when a delay or customer inquiry prompted them to investigate.
Now Wilson can reach out to individual carriers, get quotes, find the best rate and book shipping. Today, after more than a year with Cartage, Sundays is using the agent to set up API connections with the company’s different carrier partners, locating an order in the chain and predicting with better accuracy when it will be delivered.
“Instead of having to go to multiple places to guess, ‘Is it still at the middle mile? Where is it? And when’s it going to deliver?’ [Wilson] makes things much much easier,” says Sundays co-founder Moe Samieian.
Moving forward, the company is working on using the AI agent to help determine whether an order is at risk of being delayed, and at which point in the delivery process that might happen—ultimately allowing Sundays to step in, when possible, to expedite delivery. “If we can get ahead of an issue before it happens, then we can solve those issues before [they] even come to life,” says Samieian. “That is our ultimate goal.”
For Saatva—which doesn’t utilize third-party fulfillment—the benefits of hiring Wilson were primarily the 24/7 customer service offerings. The company’s origins as an online mattress disruptor meant that it was in direct competition with the bed-in-a-box brands that promised to deliver products via FedEx or UPS in 24 hours. With Wilson, the company gained clarity: It can now not only provide delivery information about a product’s whereabouts at any time of day, but also identify shipping improvements for an additional edge.
“We think we can gain efficiency with Wilson’s logistics knowledge, saying, ‘Hey, these routes are great, but what if you open a hub here? What if you open a factory here?’” says Riano. “There’s a lot we think we could scale with this, just by knowledge-gaining.”
Riano says it’s too early to tell whether or not the tool has boosted Saatva’s profitability. Similarly, for Sundays, Samieian says that success will be less about improving margins and more about improving customer satisfaction and employee well-being. With Wilson, the company has cut down on the volume of human tasks and repetitive work—so much so that staff responsibilities have changed since the AI agent was onboarded.
“We’ve been able to do more that we weren’t able to do before,” says Samieian. “We’ve been able to take [away administrative] work that a person was doing and have them do something a little bit more meaningful, a little bit less redundant.”
Basharat says this could be Wilson’s biggest asset in the home industry, a space where many founders and employees are creatively inclined yet find themselves obligated to carry out the logistics necessary to bring a vision to life. “The most important resource you have as a company is your human capital, the time of your people, and you want that to go into creative acts that require real thinking—not the operational work of emailing a trucking company, answering a customer’s question about where a shipment is,” he says. “I think logistics is one of the best applications, because then you free up your team to actually do way more interesting, important stuff.”
