Seair Exim Solutions delivers data at lightning speed with Microsoft Fabric Real-Time Analytics

Across the globe, 24,000 customers rely on Seair Exim Solutions for clutter-free, genuine export-import (EXIM) data. In 2021, the company launched Eximpedia, an easy-to-use online platform of EXIM data from 100+ countries. Today, the platform handles approximately 200,000 queries daily and is well-known for its comprehensive data sets, easy search capabilities, and bespoke visualization options.

Over the last three years, as the platform expanded, delivering on the company’s promise of superior quality, cost-effective, end-to-end EXIM data became increasingly harder. “We started Eximpedia on AWS with ElasticSearch as a search engine. It met all our requirements at the time. But as our platform grew, we were no longer able to achieve the same kind of speed and accuracy we wanted,” explains Amandeep Yadav, Co-founder of Seair Exim Solutions.

Battling with data latency

Being a user-driven application, data updates happen periodically on Eximpedia. “It’s not just about adding new data sets,” explains Yadav. “Often, a user will come to us and ask us to expand an existing data set with additional columns of data. We reached a point where every time we ran an update, it would slow down or disable the platform.” As a result, updates were often delayed to avoid disrupting normal operations. Any updates had to be scheduled overnight or on the weekend, and done only when essential.

Data lags weren’t the only issue. “We wanted to offer more complex real-time data visualization that could be customized based on each customer query. This was no longer possible with the back-end infrastructure we had in place,” explains Yadav. “We needed an urgent revamp.” The company sought a scalable solution that offered quick data updates, high data reliability, greater search flexibility, and advanced visualization.

Seair Exim Solutions turned to Mandelbulb Technologies for help. After analyzing the problems and pinpointing areas that needed the most attention, Mandelbulb came back with a recommendation—migrating from multiple service providers into one unified solution that would manage everything Eximpedia needed to function effectively.

One unified solution

Mandelbulb suggested shifting Eximpedia into the Azure Cloud and Microsoft Fabric. This would enable Eximpedia to be hosted exclusively on Eventhouses on multiple Kusto Query Language (KQL) databases and connects directly to Power BI for data and reporting. However, the migration came with some risks. “We didn’t want to create more disruptions for our users during the whole migration,” explains Yadav. “We had to limit the teething problems that tend to come up in a migration of such scale.”

The Mandelbulb team quickly went to work, devising a detailed plan for the shift. The move happened via an intermediary staging layer where all the data was verified, removing all unnecessary duplicates of data sets. The team also introduced a data lake to check data accuracy.

The migration from ElasticSearch to Eventhouse happened in just two months, with minimal interruptions and 100% data integrity rate. “Mandelbulb’s team made sure that everything was running smoothly one week after moving Eximpedia onto Microsoft Fabric,” says Yadav.

Data ingestion with no disruptions

Seair Exim Solutions saw the benefits almost immediately. “We’ve reduced operational costs by 52% by switching to a single provider to manage everything.” Today, Eximpedia has close to 10 TB of data in storage and 1 TB in production.

The feedback has been extremely favorable, explains Yadav. “While our end users might not be able to see that we’ve switched to a different solution, they have noticed a significant change in performance. We’ve received a lot of positive feedback about the variety and details available for visualization to the increase in search speeds.”

The biggest change is in the platform’s data ingestion speeds and workspace creation, which happen in real time. “Our data processing speed increased by 90%. An update that required a few hours now happens in a few minutes,” says Yadav.

Instantaneous and better search capabilities

Results come back lightning-fast now thanks to Real-Time Analytics in Microsoft Fabric. “Previously, most searches needed five to 20 seconds, depending on the type of query. Now, the average response time is under one second,” Yadav explains. And with KQL, Eximpedia users can execute more complex searches. Users can opt for full-text searches or wildcard queries that locate similar but not identical data based on a specified pattern. “The best part is that the change in speed is consistent across all these searches, no matter the complexity or depth,” he adds.

By leveraging Power BI in Microsoft Fabric, the platform offers more advanced, layered visualization options. “With Power BI, we can build charts, graphs, or bespoke visualization quickly with minimal effort,” says Yadav. Currently, 80+ dashboards are available for Eximpedia users. Customers can also opt for bespoke reporting. “Since Power BI integrates seamlessly with Eventhouse, any customer query generates a customized report in real time which is not something we were able to offer before,” he adds.

All these improvements positively impact user experience. Mandelbulb noticed a 37% increase in search activity and user interaction post-migration. “With Microsoft Fabric, we’ve been able to reshape decision-making on the Eximpedia platform,” explains Deepanshu Thakur, Co-founder of Mandelbulb. "We’ve seen a 27% annual increase in client retention and anticipate 20% growth in client acquisition over the next two years.”

It’s not only the front-end experience that improved but the back end, too. “Having everything on the same platform made it much easier for the developers to work on and scale as required,” Yadav adds. “Right out of the gate, we have strong documentation which had been missing before.”

Yadav and his team are now working to expand the platform’s capabilities by leveraging additional tools and features. “Our customers trust us because we guarantee quality, accuracy, and ease of use,” concludes Yadav. “Now that we’re back on track, we are exploring additional features such as translation services and using artificial intelligence to make the platform even more interactive for users,” he concludes.