Shipium boosts efficiency and optimizes carrier selection with MongoDB

THEIR CHALLENGE

Using MongoDB Atlas to enhance scalability and stimulate growth

As global ecommerce continues to grow, managing logistics across interconnected markets becomes essential. Retailers and logistics companies face the dual challenge of operating within thin profit margins and providing swift delivery options to keep customers happy. That’s where Shipium comes in. The company’s carrier selection software helps retailers and third-party logistics companies select the best shipping methods for their needs, whether that’s providing speedy shipping or finding the most economical option for regional deliveries.

After it was founded in 2019, Shipium originally built its logistics software on MongoDB Community Edition, the self-managed version of MongoDB’s distributed document database. Shipium chose MongoDB from the start due to its strong developer experience and the ability to scale to an enterprise level, meaning that the company could handle rapid growth when the need arose.

However, the expanding scale and complexity of Shipium’s operations have required a managed solution. “We were running MongoDB by ourselves,” said Michael Louthan, development manager at Shipium. “It was working well, but we knew that we wanted to scale and grow across multiple regions. We could have tried to navigate this internally, but we wanted a more managed path.”

To maintain the high availability of its services, Shipium needed a solution that keeps its systems operational and available without significant interruptions, meeting an uptime requirement of 99.95%. The company also needed low-latency performance across multiple geographies. Furthermore, it wanted to simplify database management by adopting a solution that included built-in best practices, such as alerts for potential issues, automatic scaling, and guardrails for compliance with international standards. After exploring potential solutions, Shipium chose to migrate to MongoDB Atlas — MongoDB’s integrated suite of cloud database and data services — in March 2023.

“MongoDB Atlas was the cleanest option for us,” said Louthan. “We didn’t need to change any of our technology; we just migrated our data over. As we dove more into the service, we discovered additional features that could help us grow. Using MongoDB Atlas, we will soon scale and transform from a single-region company to a more global enterprise.”

OUR SOLUTION

Migrating to MongoDB Atlas to coordinate multiregional logistics

In June 2023, Shipium worked with MongoDB to perform a test run of MongoDB Atlas in the company’s staging environment. After a successful test — meaning there was no reduction in performance — Shipium officially put MongoDB Atlas into production in July 2023.

“One of the things that I really like about MongoDB Atlas is I don’t need to be an expert to use it,” said Adrian Perez, DevOps engineer at Shipium. “Most of the configuration setup takes place through the user interface, and our automation scripts help provision all the infrastructure for us. When we needed technical support from MongoDB, the team was quick to respond and usually helped us the same day.”

MongoDB Atlas provides a robust, scalable database environment that handles TBs of data with minimal latency. Each day, numerous shipping decisions are processed through Shipium’s platform. Warehouses and stores send detailed order information — including destination and package specifics — to Shipium through an API. The software then processes this data in three stages: eligibility checks, date calculations, and rate comparisons.

Shipium can use smaller instances in MongoDB Atlas compared to its self-maintained MongoDB instance. Its provisioned instance size has been reduced from the equivalent of an M60 to an M40 low-CPU instance without any performance issues. With the help of the MongoDB team, Shipium also reduced its local storage usage from 325 GB to 20 GB. These optimizations help the company maximize resource efficiency while maintaining high performance.

By using artificial intelligence and machine learning (AI/ML), Shipium predicts the most efficient shipping method that will hit a specific date based on near-real-time conditions and historical data, dynamically selecting the optimal carrier for each order. MongoDB Atlas facilitates this process by hosting carrier and shipping data in one place, which Shipium can then use to run the latest AI/ML algorithms.

These capabilities are essential for near real-time decision-making, especially in an industry with seasonal fluctuations and frequently changing customer demands. “When we talk to large enterprise clients, we need to demonstrate our ability to scale to support bursts during high-volume periods, like the busy holiday season,” said Kris Gösser, chief marketing officer of Shipium. “They often ask, ‘What is your geographical coverage?’ or ‘What’s your edge computing like?’ to assess our scalability, so MongoDB Atlas helps improve our credibility to these enterprise retailers.”

OUTCOME

Achieving optimal scalability while meeting performance and latency requirements

The migration to MongoDB Atlas facilitated greater horizontal scalability for Shipium. In other words, the company can add new regions and adapt quickly to changes in demand as needed. Using this service, Shipium also enjoys high performance and low latency while maintaining consistent service delivery across various geographies. With these capabilities, Shipium can expand its business to new regions and work across different geographies.

“We have to be enterprise grade, and whatever technology that we use has to support that,” said Gösser. “Using MongoDB Atlas helps us stay competitive on latency and enables us to meet our targets.”

With MongoDB Atlas, Shipium also increased visibility into its database operations through comprehensive monitoring and metrics. Knowing that potential issues can be proactively identified and resolved without disruption has provided Shipium with a greater peace of mind. And because MongoDB Atlas is a managed service, Shipium’s developers are no longer burdened by time-consuming infrastructure management tasks like applying patches, performing upgrades, and managing backups. As a result, they can focus more on innovation and value-adding tasks. Shipium also plans to dive deeper into AI/ML to improve performance even further.

“MongoDB Atlas provides a lot of value in the realm of AI/ML,” said Gösser. “It serves as a centralized database across every single shipper and carrier. That gives us the data necessary to perform the actual data science. This is truly game-changing in the retail industry.”

To learn more, visit MongoDB Atlas.