Web app · Multi-cloud cost monitoring · Enterprise FinOps workflow.
Back to work
One shared truth
Cloud Cost Monitoring Platform
A B2B SaaS case study about turning cloud-cost overwhelm into shared visibility, earlier action, and recommendations that finance and DevOps could both understand.
40%cost savings through recommendations67%reduction in budget overruns85%weekly active usage
Case snapshot
What I owned, shaped, and proved.
UX Designer shaping research synthesis, IA, dashboards, recommendations, alerting, and handoff.
Finance leaders, DevOps teams, and cloud admins managing AWS, Azure, and GCP spend.
Stakeholder interviews, cloud-admin shadowing, competitive review, and workflow analysis.
Unified cost dashboard, AI-powered recommendations, proactive alerts, and resource drill-downs.
Create one shared cost language before adding more charts or optimization surfaces.
Design judgment
My judgment was to design one shared cost language before designing more charts. Finance needed confidence, DevOps needed specificity, and cloud admins needed control. The interface had to hold all three without making any one role feel like a guest.
12 interviews and 5 shadowing sessions showed teams building truth by stitching together 4 to 6 tools.
Cost risk was discovered after the damage was already visible in finance reviews.
A spend overview had to lead naturally into team, cluster, recommendation, and reporting actions.
Round 2 reduced anomaly identification time to 90 seconds with 100% task completion.
Process evidence
The thinking I had to make visible.
Reconstructed artifacts that show what I was trying to understand, where trust felt fragile, and how the direction became safer to build.persona map
Finance, DevOps, and cloud admin map
I used this map to stop treating finance, DevOps, and cloud admins as personas on a slide and start treating them as people negotiating the same cost reality.
Finance
Budget variance / Forecast confidence / Exportable reports
DevOps
Resource detail / Cluster behavior / Implementation risk
Cloud admins
Policy alerts / Tag hygiene / Chargeback paths
Why it matteredFinance teams saved 8 hours per month on reporting after spend data moved into one operating view.
current state map
Fragmented cloud cost picture
This map reconstructed the old reality: every role had a piece of the truth, but nobody had the whole operating picture early enough.
Finance
Monthly reviewBudget varianceSpreadsheet reconciliationLate escalation
DevOps
Resource checksCluster usageManual investigationImplementation risk
Cloud admin
Provider consolesTag cleanupPolicy alertsChargeback requests
Why it matteredTeams toggled between 4 to 6 dashboards and often discovered anomalies 5 to 7 days late.
journey
Cost prevention loop
This loop helped me frame the product as prevention, not reporting. The goal was to catch risk while action was still possible.
Detect
User signalSpend spikes were visible only after teams manually stitched reports together.
Design responseShow anomalous spend early through dashboard priority, alerts, and budget thresholds.
Explain
User signalFinance saw cost movement but not the technical reason behind it.
Design responseExpose provider, team, project, cluster, and usage drivers without forcing a console switch.
Act
User signalOptimization advice felt risky when the operational consequence was unclear.
Design responsePair each recommendation with confidence, risk, savings, and usage evidence.
Report
User signalTeams still needed proof for budget reviews and chargeback conversations.
Design responseMake savings, owner, and export paths visible so action became auditable.
Why it matteredBudget overruns decreased by 67% because teams caught spikes within 24 hours.
trust anatomy
Recommendation explainability anatomy
This artifact captured the tension inside every recommendation: saving money is only useful if the action is safe enough to take.
Current cost
The waste signal tied to a specific resource, team, or cluster.
Projected cost
The expected spend after the recommended change.
Confidence
How strongly usage patterns supported the recommendation.
Risk
Operational caveats before a team took action.
Why it matteredRecommendation engagement increased from 25% to 87% after top opportunities moved into the main dashboard with evidence.
navigation model
Spend overview to action path
This path kept users oriented as they moved from the calm overview into the technical depth where action actually happened.
01Spend overview02Team or project03Cluster or resource04Recommended action
Why it matteredTime-to-insight improved from 15 minutes to 30 seconds in the pilot.
Final design
Final Product Screens
Project artifact supplied by Savita for this case-study narrative.
Impact and results
What the work delivered.
Cost savings reached 40% through AI recommendationsBudget overruns decreased by 67%Weekly active usage increased from 18% to 85%Time-to-insight improved from 15 minutes to 30 secondsCombined annual savings reached $2.1M
Source: 12 stakeholder interviews, 400+ support tickets reviewed, and 5 cloud-admin shadowing sessions across 3 enterprise clients managing $13M+ in annual cloud spend.