Full disclosure: Spotto and Add Screenshots share a co-founder. We built Spotto because running this exact infrastructure by hand was painful — and Add Screenshots runs on it every day. This case study is what that looks like in practice, numbers included.
The Challenge: Global-Scale Cloud Complexity
Running Add Screenshots - a high-volume screenshot API serving customers from US, Europe, and Australia / Asia Pacific regions—requires a sophisticated Azure infrastructure. Our architecture spans multiple continents and includes:
- Azure Serverless Functions: Processing millions of screenshot requests with automatic scaling based on demand.
- Azure Container Apps: Running headless browser instances for rendering complex web pages.
- Virtual Machines: Dedicated compute for high-throughput GPU supported regions.
- Cloudflare WAF & CDN: Protecting our API endpoints and caching responses at the edge globally.
- Azure Blob Storage: Holding operational data such as logs, usage reports, and short-lived opt-in CDN cache artifacts — never customer screenshots, which are delivered directly to customer-owned storage.
As our user base grew globally, managing this infrastructure became increasingly complex. We faced several challenges:
- Rising Costs: Tracking "zombie" resources - unused disks, over-provisioned resources, and unoptimized storage tiers - that were inflating our monthly bill across three regions.
- Scaling Inefficiencies: Ensuring our Container Apps scaled appropriately without over-provisioning during off-peak hours.
- Security Drift: Maintaining consistent WAF rules, TLS policies, and access controls across all deployments.
- Billing Complexity: Understanding cross-region data egress costs and optimizing reserved instance utilization.
The Solution: Spotto's AI-Powered Cloud Optimization
We integrated Spotto, an AI-powered cloud optimization platform, into our workflow. Spotto acts as an "always-on" cloud architect that continuously scans our Azure subscriptions for inefficiencies, misconfigurations, and security risks.
What sets Spotto apart is its approach to intelligent cloud operations. As per their documentation, Spotto provides:
- AI-Powered Anomaly Detection: Automatically identifies unusual patterns in resource usage and costs.
- Proven Fix Recommendations: Suggests remediations based on best practices with context, effort estimates, and risk levels.
- Autonomous Healing Capabilities: Enables high-confidence fixes across Azure infrastructure.
Key tool: the Spotto Autopilot browser extension. The game-changer for our team was the Spotto Autopilot browser extension. Instead of logging into a separate dashboard, our engineers see cost insights and recommendations directly inside the Azure Portal. When viewing a VM, Container App, or Storage Account, Spotto overlays actionable insights right where we take action.
How We Onboarded Spotto
Setting up Spotto was straightforward and non-invasive:
- Secure Read-Only Access: We registered an Azure AD (Entra ID) Application with Reader and Billing Reader roles. Spotto only reads metadata—it never writes or manages secrets.
- Storage Inventory Analysis: We generated Azure Blob Inventory reports with Last Access Time tracking and uploaded them to Spotto for storage tiering recommendations.
- Continuous Scanning: Spotto now runs multiple daily scans across all three regions, alerting us to misconfigurations as they arise.
Optimization Results by Resource Type
Serverless Functions & Container Apps
Managing serverless workloads across three regions meant dealing with cold starts, regional latency variations, and scaling policies that didn't always match actual demand. Spotto helped us:
- Identify over-provisioned app service plans that were allocated premium plans but rarely exceeded consumption tier limits.
- Optimize Container App scaling rules based on actual usage patterns rather than worst-case estimates.
- Right-size our region deployments to match traffic patterns by time zone.
Virtual Machine Rightsizing
Our regions run dedicated multi-CPU VMs — at least four per region, 12+ across the platform — for high-throughput screenshot rendering. Spotto's VM analysis revealed:
- Unattached Disks: Premium SSDs left behind after VM deletions across multiple regions.
- SKU Mismatches: Development environments running on production-grade SKUs during off-hours.
- Idle Resources: VMs in lower-traffic regions that could be scaled down during off-peak hours.
Web Application Firewall & CDN Migration
Maintaining consistent WAF policies across our regional Azure deployments was a manual nightmare—and expensive. Spotto's cost analysis helped us realize we could save significantly by migrating to Cloudflare for both our WAF and CDN. The results:
- Consolidated WAF management into a single Cloudflare dashboard instead of managing Azure WAF instances per region.
- Reduced latency with Cloudflare's global edge network caching our API responses closer to end users.
- Simplified security policies with consistent rules applied globally rather than per-region configuration drift.
- Achieved substantial cost savings compared to Azure Front Door and regional WAF deployments.
Storage Tier Optimization
Operational data — logs, usage records, and diagnostics — made storage a significant line item. Spotto's blob storage analysis helped us achieve 60% savings by:
- Moving older logs and usage records to Cool and Archive tiers based on access patterns.
- Implementing lifecycle policies to automatically transition data between tiers.
- Identifying redundant data stored across multiple regions unnecessarily.
Billing & Scaling Insights
Understanding our Azure bill across three regions was challenging. Spotto provided:
- Clear visibility into cross-region data egress costs.
- Reserved instance utilization analysis to maximize savings.
- Alerts for unexpected cost spikes in any region.
Security Improvements
Beyond cost optimization, Spotto strengthened our security posture without blocking legitimate traffic:
- Validated that all Storage Accounts enforced "Secure Transfer Required" across all regions.
- Identified RBAC assignments that were overly permissive for third-party integrations.
- Flagged inconsistent security policies between regional deployments.
- Confirmed our Cloudflare Tunnel configuration keeps VMs off the public internet entirely.
Improved Developer Velocity
Spotto offers multiple integration options that fit different workflows:
- MCP Integration: Connect AI assistants like Claude directly to your Azure data via Model Context Protocol for instant infrastructure questions.
- Browser Extension: Get real-time cost insights and recommendations directly in the Azure Portal.
- REST API: Integrate Spotto into your CI/CD pipelines and automation workflows.
Our developers now spend less time debugging configuration issues and more time building features. The AI assistant answers infrastructure questions instantly, acting as a force multiplier for our DevOps team.
Conclusion: A Strategic Partner for Global-Scale Azure Infrastructure
For Add Screenshots, Spotto isn't just a monitoring tool - it's a strategic partner that keeps our infrastructure lean and secure across US, Europe, and Australia / Asia Pacific regions. Whether you're managing Serverless Functions, Container Apps, Virtual Machines, or complex WAF configurations, Spotto's AI-powered insights help you identify misconfigurations and optimize costs with simple, actionable recommendations.
We can confidently scale our screenshot API knowing that Spotto is continuously watching our back—detecting anomalies, recommending proven fixes, and helping us maintain a cost-effective, reliable service for our global customer base.
Try Spotto for free
See what your Azure subscriptions can save in minutes — secure read-only access, revoke anytime. No credit card required.
Need global screenshot coverage?
Add Screenshots captures from the US, Europe, and Australia / Asia Pacific on 12+ dedicated servers with automatic geo-failover and no concurrency caps. Free tier: 100 screenshots every month.