How Uni streamlined their cloud infrastructure costs by 20% with Cloud Cost Observability
Year
Uni is a new-age fintech company with a mission to change the way people interact with credit. By reimagining the existing credit products and delivering an unmatched credit experience to millions of Indians, Uni is disrupting how traditional credit products work.
Industry
Region
Key Features
Cloud Optimization, K8s Optimization
Introduction
The credit card industry in India has grown significantly, with innovations in technology leading to smooth onboarding, personalized offers, and better mobile apps, benefiting and attracting customers. The industry achieved a CAGR of 20% over the past five years, with credit cards exceeding 78 million in July 2022. The rapid scale and growth of this industry is attributed to many factors with technology being a key among them.
Challenges Faced
Uni is a fast emerging player in this industry and one of the biggest requirements to stay competitive was to keep their technology and infrastructure costs under consistent check. The team were actively working on staying ahead of escalating costs, ensuring optimal resource utilization and driving a strategy towards effective cost management. Along with these efforts, Uni also wanted a system that helped their teams identify suboptimal instance types, oversized Kubernetes nodes and pods, and underutilized Savings Plans and Reserved Instances (RIs).
“Staying competitive in the financial industry, requires being able to understand customer needs, usage patterns and a wide range of other parameters quickly and consistently. Tech costs are continuously monitored and maintaining lower cloud spend is a growing priority for teams.”
Abhra Dasgupta
Uni
Solution
The team at Uni was introduced to Cloud Cost Observability through Amnic and were given pilot access to use the Amnic platform. Connecting to Amnic proved easier than anticipated owing to stringent security protocols set in place by the platform including SOC 2 Type 2 compliance, GDPR and ISO: 27001 certifications. In 24 hours, Uni was able to gain a 360-degree view into their infrastructure and costs.
By using Amnic, Uni was able to gain visibility into costs associated with various resources, instances, as well as efficiency of their nodes and pods. This allowed the team to take action where necessary and implement measures that would enable them to run a lean cloud infrastructure. Some of the key implementations driven by Amnic included
Optimizing Amazon CloudWatch: By fine-tuning Amazon CloudWatch configurations, Uni reduced unnecessary monitoring costs while retaining essential performance insights.
Optimal Instance Type Recommendations for EC2: Leveraging data analytics, the partner provided recommendations for optimal EC2 instance types, ensuring the best performance-to-cost ratio based on Uni workload requirements.
Right-Sizing of Kubernetes Nodes and Pods: A thorough analysis of Kubernetes usage identified opportunities to right-size nodes and pods, aligning resources more closely with actual demand and avoiding over-provisioning.
Results
20% Lower cloud infrastructure costs
By using Amnic, Uni was able to instrument a system of cloud cost observability that affected a comprehensive series of changes. This provided them with the knowledge base and resources to manage an effort of cost optimization and improved engineering efficiency. Acting on the insights from Amnic led to a 20% reduction in cloud infrastructure expenses, within the first month of implementation.
“Working with Amnic has paved the way for a journey where we have control and comprehension of the overall cloud spend. While we have been able to successfully optimize our infrastructure, Amnic is playing the long term role of watching out for unprecedented costs and giving us insights into spends that could be better managed over time.”
Abhra Dasgupta
Uni
Uni's partnership with Amnic proved instrumental in addressing their cloud cost challenges. By optimizing Amazon CloudWatch, selecting optimal EC2 instance types, right-sizing Kubernetes nodes and pods, and leveraging Savings Plans and Reserved Instances, Uni significantly reduced their cloud spending while maintaining high performance and scalability. This success is a demonstration of the growing need for cloud cost observability within the financial sector.
