🧾 Slide 1 — Cloud Service Comparison (Core Differences)

What are we really comparing?

Not vendors — levels of responsibility and operational ownership

Baseline Assumption (Same for all):

  • Windows-based backend (Delphi API)
  • React frontend (managed hosting + CDN)
  • Managed SQL Server (PaaS, not VM)
  • Built-in backup, High Availibility, monitoring, security
  • Enterprise SLA expectations

🔹 Microsoft Azure

(PaaS-first Microsoft-native platform)

Equivalent Services:

  • App Service (Windows) – Delphi API
  • Azure SQL (Managed) – MS-SQL
  • Static Web Apps / Front Door – React
  • Azure Monitor, Backup, Defender

What you get:

  • End-to-end managed platform
  • Native MS-SQL & Windows optimization
  • Built-in HA, backups, patching
  • Strong enterprise SLA Service Level Agreement
Strength: ✅ Lowest operational risk
✅ Best fit for Microsoft workloads
Trade-off: ⚠ Higher baseline cost
⚠ Less infra-level customization

🔹 Amazon Web Services

(PaaS via managed services, not default)

Equivalent Services:

  • Elastic Beanstalk (Windows) – Delphi API
  • RDS SQL Server – Managed MS-SQL
  • S3 + CloudFront – React
  • CloudWatch, Backup, Shield

What you get:

  • Same enterprise-grade capabilities
  • Very mature cloud ecosystem
  • Strong global scale
Strength: ✅ Maximum flexibility
✅ Best multi-cloud & hybrid support
Trade-off: ⚠ Higher ops effort than Azure
⚠ Cost grows fast if not tightly governed

🔹 Google Cloud Platform

(Cloud-native PaaS with Windows as secondary)

Equivalent Services:

  • Compute Engine (Windows) – Delphi API
  • SQL Server on Compute Engine
  • Cloud Storage + CDN – React
  • Cloud Monitoring & Logging

What you get:

  • Excellent performance & networking
  • Clean developer experience
  • Strong AI/data tooling
Strength: ✅ Best analytics & AI ecosystem
✅ Competitive compute pricing
Trade-off: ⚠ No true managed SQL Server PaaS
⚠ Weak alignment with MS stack

🔹 Netmagic (NTT)

(Managed hosting, not cloud-native PaaS)

Equivalent Services:

  • Managed Windows Server
  • SQL Server on VM
  • CDN / Hosting via provider
  • Manual monitoring & backups

What you get:

  • Fully outsourced operations
  • Strong India compliance & support
  • Fixed-cost predictability
Strength: ✅ Minimal internal DevOps needed
✅ Local compliance & support
Trade-off: ⚠ Not cloud-native
⚠ Slower scaling & innovation

🔹 VPS
(Self-Managed)

Equivalent Services:

  • Windows VM Machine – Delphi API
  • SQL (Self Managed) – SQL-SMS
  • IIS setup | Xampp setup for hosting – React
  • Self Monitor, Self Backup, Microsoft Defender

What you get:

  • Full control over OS, apps, DB
  • All responsibility lies with engineering team – Self Monitor, Self backups, patching
Strength: ✅ Low infrastructure cost
✅ High control
Trade-off: ⚠ Higher Operational cost
⚠ High operational risk
1 / 2

🧾 Slide 2 — Cost, Risk & Recommendation

💰 Cost Reality (Monthly Perspective)

Platform Infra Cost Hidden Cost
Azure Medium (£1.2k) Low
AWS Medium–High Medium
GCP Medium Medium
Netmagic High Low
VPS Low High (engineering time, outages)
VPS looks cheap because reliability, security & monitoring are not included

⚠️ Operational Risk Comparison

Area Azure AWS VPS
Backup & DR Managed Managed Manual
Monitoring & Alerts Built-in Built-in DIY
Security Platform-level Platform-level DIY
Outage Handling Auto-heal Manual Manual
SLA Yes Yes No

🎯 Key Insight (Core Difference)

Cloud platforms reduce operational responsibility;
VPS shifts responsibility to engineers.

✅ Recommended Strategy

Hybrid, Cost-Aware Approach

  • Production: Azure (use existing credits & MS ecosystem)
  • Dev / QA / Experiments: VPS
  • Cost Optimization: Tune Azure services instead of migrating

🧠 Final Decision Statement (Use This Verbatim)

Azure costs more because it absorbs operational risk.
VPS saves money by transferring that risk to the engineering team.

"We should optimize cloud usage before replacing managed reliability with manual responsibility."

2 / 2