For more than two decades, Jomashop has built a reputation as a leading online retailer for luxury watches, fragrances, handbags, and accessories. “We’re a B2C company selling everything from $50 items to half-million-dollar watches,” says Isaac Montaine, Chief Product Officer at Jomashop. “We’re based in New York, but our reach is global.”

With over 250,000 products listed on its site and hundreds of thousands more passing through its warehouse each year, Jomashop’s success depends on keeping its vast catalog accurate, searchable, and up to date. “We sell in every category, every price point, and every payment method—from credit cards to Bitcoin,” Montaine says. “To make that work, our systems need to be fast and clean.”

The Challenge: Managing Product Data at Scale

As Jomashop expanded from watches into categories like fragrances, its team encountered a major obstacle: inconsistent, messy product data from multiple suppliers and drop shippers. “Every manufacturer describes things differently,” Montaine explains. “You’ll have one saying ‘tea tree,’ another saying ‘tea-tree oil,’ and others throwing in extra marketing terms that mean nothing to the shopper.”

Before adopting AI, Jomashop’s team spent extensive time cleaning and labeling data manually—an unsustainable process for a catalog that grows daily. “When you have 20,000 perfumes, you can’t go one by one,” says Montaine. “We needed automation to evaluate and describe products quickly so we could get them to market faster.”

The Solution: DigitalOcean and DigitalOcean AI Platform

Jomashop began experimenting with automation long before the current wave of AI tools. “Even before AI was a thing, we were trying to automate product descriptions,” Montaine recalls. Initially, Jomashop used DigitalOcean Droplets and Functions to host APIs that connected to OpenAI’s models, enabling automatic product data enrichment.

“When we started, setting up AI workflows took days,” says Montaine. “We used DigitalOcean Functions to call the API, return results, and plug them back into our product feed. It worked—but it took a lot of time to maintain.”

That changed when Jomashop adopted DigitalOcean AI Platform. “Now, we can build and deploy an agent in minutes,” Montaine says. “Instead of spending 70% of our time on infrastructure, we can focus entirely on improving accuracy.”

Today, Jomashop runs multiple internal agents on the DigitalOcean AI Platform to enhance product descriptions, standardize naming conventions, and categorize items. “Our agents are internal—focused on product data, SEO, and site structure,” Montaine explains. “They clean up information before it reaches the customer.”

The company uses DigitalOcean Droplets and Managed Databases (including MongoDB) to support these AI agents. “We use a MongoDB database on DigitalOcean for our product feeds,” Montaine says. “It reduces stress on our main servers and gives developers a centralized, reliable data layer to work with.”

The Impact: Faster Iteration and Stronger SEO

Since adopting the DigitalOcean AI Platform, Jomashop has dramatically improved its ability to iterate quickly, keep data clean, and publish accurate listings at scale. “The platform makes it easy for non-technical users like marketers to test and improve results,” Montaine says. “It looks familiar—like ChatGPT—so they can jump right in.”

Beyond operational efficiency, the business has seen tangible SEO benefits. “When we cleaned and standardized our data, it actually improved our Google rankings,” Montaine explains. “Before, messy data meant duplicate or conflicting content that cannibalized our search performance. Clean, structured data means one clear result that ranks higher and drives more traffic.”

The Future: Expanding Agent Capabilities

Looking ahead, Jomashop plans to migrate more of its existing AI workflows to the DigitalOcean AI Platform for simplicity and scalability. “We have agents running on other platforms like Vertex AI, but it’s harder to maintain,” Montaine says. “We’ll move them to DigitalOcean AI Platform so we can iterate faster and give them access to shared knowledge bases.”

Ultimately, Jomashop sees AI as a way to expand its bandwidth and accelerate innovation. “If we wanted to build an entire new category—say, grouping perfumes by fragrance family—it would take a human team half a year,” Montaine says. “With AI, we can set it up, let it run, and iterate on the accuracy of it.”

For Montaine, DigitalOcean provides the right balance of power, simplicity, and affordability. “The easiest way to start building was with DigitalOcean,” he says. “It’s reliable, easy to scale, and gives us the flexibility to focus on results instead of infrastructure.”