Skello scales seamlessly with MongoDB Atlas

THE CHALLENGE

Building resilience for millions of transactions

In any industry, managing a workforce is complex—involving scheduling, compliance hurdles, payroll logistics, and clear communication. As companies strive for efficiency amid such demands, a robust digital solution becomes essential. Skello answers that call. Founded in France in 2016, this fast-growing SaaS provider specializes in HR management solutions for field employees in sectors such as hospitality, retail, healthcare, industry, and services. Today, the company has nearly 400 employees, including a strong development team, and serves 22,000 enterprise clients with 600,000 daily users.

Skello’s recent transition from nimble startup to ambitious scale-up revealed the limits of its legacy architecture. As user volumes soared and its client base diversified, the patchwork of systems that once served it well began to strain. “What was built in the early stages of Skello was very good at the time and helped the company grow,” says Cédric Peruzzi, one of Skello’s’s two technical architects. “But now we need to make sure that growth can continue in the future.”

Processing millions of daily HR transactions—shift swaps, payroll runs, compliance checks, and notifications—required both resiliency and cost control. Every technical decision, from infrastructure reviews to the design of new features, had to serve Skello’s business vision: empowering managers and frontline teams with intuitive, high-performing tools that could grow with their needs. The company’s architectural rethink was also about fostering internal agility, enabling developers to own their solutions while staying aligned with platform goals.

To keep pace with customer demands and support millions of daily interactions seamlessly, Skello needed a database that matched its ambition for speed, scale, and adaptability. Enter MongoDB, providing the foundation for the next phase of the company’s growth— a data engine built for resilience, flexibility, and innovation.

OUR SOLUTION

A flexible AI-ready architecture for 600K users

Skello’s technology landscape is built to balance efficiency, scalability, and innovation. The platform runs on a 50/50 split between Ruby and Node.js, hosted on AWS serverless infrastructure for seamless scaling. A microservices architecture underpins the system, supported by databases tailored to different needs: PostgreSQL, DynamoDB, and, most recently, MongoDB.

Peruzzi and his team first explored MongoDB to enable offline functionality—a lightweight replica stored on customers’ smartphones that syncs once connectivity resumes. From there, they found MongoDB offered a sweet spot between a raw key-value database and a traditional SQL database. This flexibility proved ideal for Skello’s AI-driven needs, particularly vector databases that support efficient search and retrieval. Other options fell short. PostgreSQL required cumbersome server updates, and AWS OpenSearch was prohibitively expensive at the smaller data volumes Skello managed on its previous architecture. “Now we have a crazy amount of data,” explains Peruzzi. “But that's OK—MongoDB suits our needs.”

With MongoDB Atlas Search, Skello gains an embedded search engine that handles fuzzy search, accommodating typos and natural language queries. “It ticks a lot of boxes on our side,” says Peruzzi. “It's efficient, it works very well, we have a really good response time, and we don't have to worry about the synchronization of the search engine versus the database—MongoDB handles all that stuff.”

Today, Skello uses MongoDB Atlas to power key capabilities, including an AI-enabled chatbot that allows managers to query employee data such as probation periods, attendance, termination dates, or gender hour allocation. MongoDB Atlas also supports a document management feature crucial for payroll mandates, securely uploading and distributing payslips to employees’ private systems. Here, AI combined with Optical Character Recognition (OCR) in MongoDB processes diverse payslip formats, especially from international providers. RAG enhances Skello’s document processing by accurately identifying problematic providers and verifying document types, ensuring only payslips are processed.

With MongoDB’s robust, flexible database infrastructure, Skello can innovate quickly, scale reliably, and continue enhancing user experiences as it grows.

OUTCOME

Powering growth through performance and efficiency

For Skello, the impact of MongoDB has been transformative—both for customers and  its development teams. By powering critical features like AI-enabled search, payroll document processing, and conversational chatbots, MongoDB Atlas has enabled the company to scale confidently while delivering tangible value to clients.

One clear success has been Skello’s payslip management feature. Every month, the platform processes around 100,000 documents, with huge spikes in activity during the first few days of the month when payroll is distributed. Using RAG has boosted performance, raising retrieval rates for previously low-performing documents to 70–80%. Even at peaks, MongoDB Atlas handles the load without performance issues. “We went from 40,000 pages when we released the feature in October last year to 100,000 in May this year,” explains Peruzzi. “Customers love it, and we’ve never had an impact on performance.” The feature has proven so reliable that businesses now pay for it as a premium add-on, saving their HR teams hours of manual effort each month.

Beyond database performance, MongoDB Atlas has modernized the developer experience at Skello. Most engineers were new to the database, but quickly got up to speed thanks to high-quality documentation and MongoDB University’s focused, hands-on training. “MongoDB has really good documentation,” says Peruzzi. “If everything could have that level of quality, we would be swimming in happiness!” Within a week of being introduced to MongoDB, developers were delivering business value. For Skello, it means faster onboarding, quicker time-to-market for new features, and a development culture empowered to experiment and innovate.

Cost efficiency has also been a standout. “The ratio of cost versus benefits is one of the best I’ve seen so far,” says Peruzzi. With even the smallest MongoDB Atlas cluster, Skello can handle “multiple millions of documents” while offloading upgrades, backups, and scaling to MongoDB. “It just works,” adds Peruzzi. “When we have a problem, we know it’s on our side—not on MongoDB’s.”

Perhaps most importantly, MongoDB has opened new doors. “Everything AI-related is enabled by MongoDB,” says Peruzzi, from RAG to embedded search. Looking ahead, Skello plans to migrate additional services to MongoDB, simplifying architecture, reducing code complexity, and positioning itself to scale in the next chapter of growth.