How ClickIT Cut Checkout Latency by 75% and AWS Costs for Bigger Picture

Services

DevOps

Industry

Entertainment Technology

We modernized their cloud infrastructure with advanced SRE practices, reducing latency and cutting operational costs

Bigger Picture is an intuitive, cloud-based management platform providing end-to-end operational solutions.

ClickIT’s SRE team built auto-scaling caching solutions. They streamlined database workloads. They set up end-to-end observability. They optimized resource use across their AWS ecosystem.

The Strategic Decision at Stake

Without stabilizing cloud performance and resolving resource bottlenecks during peak traffic hours:

Slow checkout flows and 504 gateway timeouts during ticket drops would lead to immediate lost revenue and poor user experience
Unmanaged database connections would risk total system outages during major venue events
Unoptimized Redis clusters and runaway logging policies would continue driving unnecessary monthly cloud spend
Silent failures would go undetected due to noise-heavy, misconfigured alerting systems

The Challenges Bigger Picture Faced

Bigger Picture needed to stabilize its event management infrastructure to reliably handle heavy user traffic without incurring unsustainable operational expenses.

Key challenges included:

Database Bottlenecks: SQL Server reaching resource limits during peak hours due to uncleaned connections.
Costly Cache Bottlenecks: Redis memory scaling reaching limits with high operational overhead.
Limited System Observability: Ineffective alerting frameworks and unmanaged log retention driving up AWS CloudWatch costs.
Transaction Latency: Checkout transaction times taking up to 60 seconds during peak evening booking hours.
Resource Inefficiencies: Unused infrastructure and cross-region transfer fees inflating monthly AWS bills.

Our Approach: Site Reliability Engineering & Cost Governance

To resolve operational bottlenecks and build a resilient infrastructure, ClickIT’s team implemented a targeted SRE strategy:

Database Optimization & Connection Hygiene

Configured automated connection cleanups and integrated real-time Slack notifications via AWS CloudWatch and Performance Insights to proactively detect and mitigate high database load.

Dynamic Auto-Scaling for Caching

Engineered an automated scaling framework using AWS Lambda and CloudWatch metrics to dynamically scale cache nodes up or down based on real-time traffic spikes, optimizing both performance and cost.

Enhanced Observability & Log Management

Standardized monitoring by deploying Fluent Bit, Loki, Grafana, and Prometheus.

Overhauled alarm policies, cleared irrelevant noise, and configured strict log retention and deletion rules in OpenSearch to prevent silent failures.

Workload Distribution & Latency Reduction

Restructured workload distribution across Kubernetes pods managed via Karpenter and Helm to eliminate bottlenecks during checkout.

Cloud Cost Optimization

Audited AWS environments using AWS Cost Explorer, removed idle resources and unfinished POCs, and eliminated cross-region data transfers to reduce infrastructure overhead.

This comprehensive approach restored platform reliability during high-demand events while establishing long-term cost efficiency.

All Technologies Used

We implemented a robust tech stack for cloud reliability, automated scaling, observability, and cost governance.

AWS

EKS

RDS

ElastiCache

CloudFront

CloudWatch

Lambda

Kubernetes Open Search

Karpenter

Grafana

The Strategic Outcome

The Results for Bigger Picture

As a result of this work, the client cut Redis bandwidth use by 55%.
They also reduced RDS database connections by 75%.
This led to major cost savings. AWS CloudWatch costs dropped from about $1,300 to about $499 per month.
Our engineers improved the end-user experience. They reduced peak evening web transaction times from 60 seconds to 8 to 14 seconds.
The number of 504 timeout errors during checkout dropped a lot. The new observability setup made alerts work well and prevented silent failures.

Decisions Unlocked

Confidently support massive live ticketing events and peak venue booking surges without service degradation
Reallocate engineering bandwidth from troubleshooting infrastructure outages to building core SaaS product features
Maintain predictable cloud infrastructure costs with clear resource allocation rules