Uni Cards

From OSS to Grafana Cloud: How Uni Cards streamlined observability and freed up engineering time

When Uni Cards, a fast-growing fintech startup based in India, needed to modernize its observability setup and reduce its operational burden, the Uni engineering team knew just where to turn. After years of using other vendors and Grafana OSS, the team migrated to Grafana Cloud, which not only lowered their overhead but also cut their metrics costs by roughly $12,000 per month.

“Grafana is one of the best tools in the industry, so it was an obvious choice for us,”  said Rahul Golani, Principal Engineer at Uni.

Uni partners with leading banks to offer co-branded credit cards, with a standout rewards program that gives customers digital gold every time they swipe. Users can also purchase and lease digital gold directly or redeem rewards as popular shopping vouchers.

Rahul has been at Uni Cards for over four years, working across both the credit card and gold rewards platforms. His lean engineering team is laser-focused on shipping high-quality features quickly and Grafana Cloud has given them the reliable analytics and observability foundation they need to do just that.

Rahul recently spoke with Grafana Labs about how migrating to Grafana Cloud helped Uni Cards simplify their observability stack, free up valuable engineering time, and cut costs.

What led your team to adopt Grafana Cloud, and what challenges were you trying to solve?

We were using the ELK stack for most of our observability needs, primarily for logs, and to some extent for traces. But ELK didn’t really support tracing natively, and we needed something more powerful with more functionality. That’s what initially led us to try Grafana OSS.

We ran Grafana OSS for about two years. It gave us much better tracing capabilities, especially with Tempo, which let us generate real traces and create metrics from them. That was a big improvement over what we had with ELK. It was also more cost-effective, because ELK had become quite expensive for the functionality it provided. Grafana OSS was both cheaper and more capable, so the move made sense.

Eventually, as our projects grew, maintaining the OSS stack became challenging. We were spending a lot of engineering effort on upgrades, security patches, scaling the infrastructure, and managing alerts. Any time something broke or we needed to scale Grafana or Prometheus, it required internal expertise and hands-on work.

That’s when we started looking for a managed solution. Grafana Cloud allowed us to offload all of that operational burden like updates, security, scaling, while giving us the same functionality. When we factored in not just the software costs but also the cost of the people managing the system, Grafana Cloud was the clear choice. It gave us expert-managed infrastructure and let our team focus on building instead of maintaining.

How has Grafana Cloud helped reduce operational overhead and free up engineers to innovate?

At one point, Uni Cards had a fairly large engineering team, but later our organization split into three separate companies. After that, our team became much smaller. That was a key reason why moving to Grafana Cloud made sense—it allowed us to offload a lot of operational work. With a smaller team, it was much easier to “ship it to the cloud and forget about it,” instead of constantly managing and maintaining the stack ourselves.

Transitioning from OSS to cloud required a short adjustment period for cost optimization, but after three months we stabilized and found the right balance.

Since then, the difference has been huge. The Kubernetes dashboards and visualizations in Grafana Cloud are much more intuitive than what we had before. Scaling is now automatic, we can create as many alerts as we need, and the system just handles it. We’ve set up alerts for every API and key service, so if anything goes wrong, we know immediately.

Most importantly, we no longer have to worry about upgrades, security patches, or compliance issues. In the finance domain, audits are frequent and every tool must be up to date and compliant. With OSS, that responsibility was on us. Now, Grafana Cloud handles it, which frees up our engineers to focus on innovation instead of infrastructure.

We’ve also been able to optimize our observability footprint, reducing unused metrics and trimming log volume using the built-in dashboards that show which services are pushing the most data. Those capabilities helped lower costs and made our observability much more efficient.

Tell me about your experience with Adaptive Metrics. 

Adaptive Metrics has been really nice. Billing was a big topic for us, and when I joined no one had tried using Adaptive Metrics yet. I decided to give it a try and honestly, it was one of the easiest and most impactful things I’ve done.

The system automatically applied most of the recommended rules, and I only had to create two or three custom rules myself. In total, it ended up applying around 500 to 600 rules. That alone cut our metrics volume from about 4 million down to 1 million. With a bit more fine-tuning, I reduced it further to around 700,000 metrics.

In terms of cost, previously, we were paying around $12,000 per month just for the additional 3 million metrics beyond our 1 million metric allowance in Grafana Cloud. By reducing our usage down to around 700,000 metrics, we’ve essentially eliminated that extra cost.

For us, that has been huge. It not only brought our costs under control, but it also let us keep the system lean without sacrificing visibility.

Now that you’ve been using Grafana Cloud for some time, how has the overall experience been?

One of the big reasons we chose Grafana Cloud was that our team was already very familiar with Grafana from using the OSS version. That made the transition smooth,  there wasn’t a big learning curve or change management effort. It just made sense to stick with a tool that everyone already knew and trusted, rather than switch to something new for only a small potential gain. 

Overall, the product itself has been excellent. I can confidently say Grafana Cloud has been a great fit for us.

Uni Cards logo
Industry
Financial Services
Company Size
100+ employees
Headquarters
Bangalore, India
82%
decrease in metric volume with Adaptive Metrics
$12K
a month savings thanks to Adaptive Metrics