Former Bolt CEO Maju Kuruvilla’s new startup just cleared a big milestone.
Spangle, his AI-powered e-commerce company, has raised a $15 million Series A at a $100 million post-money valuation, the company said. The all-equity round was led by NewRoad Capital Partners, with Madrona, DNX Ventures, Streamlined Ventures and several strategic angels joining in.
It’s a sharp jump from Spangle’s $6 million seed round, which valued the Seattle-based startup at a $30 million pre-money. In total, Spangle has now raised $21 million.
Betting on AI-native shopping
Retailers are quietly facing a big shift: more product discovery now happens before a shopper ever lands on a brand’s website. AI assistants, social feeds and recommendation engines increasingly steer what people see and buy.
Kuruvilla wants Spangle to be the layer that connects that messy discovery world to the actual storefront.
Instead of sending visitors to fixed product or category pages, brands using Spangle route traffic to what is essentially a blank canvas. Spangle’s AI then builds the page in real time.
At the center of this is a proprietary model Spangle calls ProductGPT. It pulls in signals like:
- Where the shopper came from
- What they searched for or clicked on
- How similar visitors have behaved
Using that data, ProductGPT decides which products, recommendations and content to show in that moment.
“We are future-proofing the brand,” Kuruvilla said, explaining that Spangle trains its AI model on each retailer’s catalog and performance data so the experience can adapt automatically.
Some observers already describe Spangle as a kind of Shopify for AI-powered commerce infrastructure rather than just another optimization tool.
Early traction with fashion brands
Spangle only emerged from stealth in March 2025. Since then, it has signed nine enterprise customers, including fashion names Revolve, Alexander Wang and Steve Madden. According to Kuruvilla, those three alone drive about $3.8 billion in combined online sales.
Traffic flowing through Spangle’s platform is growing fast. The company says it has seen roughly 57% month-on-month growth in traffic, with every customer expanding how they use the software.
While Spangle doesn’t share hard revenue numbers, it says it quadrupled its annualized revenue in the fourth quarter.
For brands, the pitch is straightforward: more money from the same traffic.
Kuruvilla said retailers using Spangle are seeing:
- Close to a 50% increase in revenue per visit
- A doubling of return on ad spend (ROAS)
- Around a 15% increase in average order value (AOV)
Revolve is one of the early case studies. Spangle’s software has helped the retailer adapt shopping experiences in real time, driving about a 60% improvement in ROAS and a 50% increase in revenue per visit, according to Ryan Pabelona, Revolve’s vice president of performance marketing.
Built for a world of AI agents
Spangle’s timing lines up with a broader shift in how people shop.
More consumers now rely on AI tools like OpenAI’s ChatGPT and browser-based agents to research products, compare options and even start the checkout process. In that world, a brand isn’t just serving human visitors—it’s also serving machines that are effectively shopping on behalf of users.
Kuruvilla argues that traditional static product listing pages can’t keep up with that change. Brands need software that can respond dynamically to both humans and bots hitting their sites.
He said Spangle only became viable in the past two years, as three trends converged:
- Consumers became comfortable discovering products through AI tools.
- Discovery channels exploded beyond Google and Meta, fragmenting the customer journey.
- Advances in AI slashed the cost and latency of generating real-time, personalized experiences.
Together, those shifts made it possible to move from incremental fixes—like A/B tests on a homepage—to an “AI-native” commerce stack that rewrites the storefront on the fly.
From Amazon and Bolt to a new infra layer
Kuruvilla brings a heavyweight commerce resume to Spangle. Before founding the startup in 2024, he was CEO of one-click checkout company Bolt. Earlier, he spent more than a decade at Amazon working on large-scale commerce and AI systems.
He started Spangle with CTO Fei Wang, a former Amazon principal engineer who worked on Alexa and customer service technologies and later became CTO at Saks Off 5th.
Their shared background in running big commerce and payments platforms shaped how they see the opportunity. Rather than bolt-on tools, they want to build underlying infrastructure that retailers plug into.
A $100M company with six employees
One detail stands out in Spangle’s numbers: the entire company currently has six full-time employees.
That headcount underlines a broader theme across AI startups. Modern AI tooling—from model hosting to code generation—is allowing small teams to build and run enterprise-grade software with far fewer people than a decade ago.
The new capital will not stay idle. Kuruvilla said Spangle plans to invest heavily in research and development, grow its engineering team and build out a dedicated sales organization to go after more large retailers.
The race is on to define what an AI-native storefront looks like. With $100 million on the board and marquee fashion brands already in production, Spangle now has the cash—and the pressure—to prove its ProductGPT approach can scale far beyond its first nine customers.
