iTransform · Production hosting

Hosting Cost Explorer

Interactive AWS cost model for a self-hosted iTransform production deployment, built on the managed-container architecture the platform already runs on today. Move the sliders to size it for your organization's reach.

us-east-1 pricing, pulled 24 Aug 2026 Source: AWS Pricing API + published rate cards Production single environment Capacity constant validated by load test, 24 Aug 2026
Recommended · AWS / month
—
Managed containers, on-demand pricing
Cost per active user
—
per MAU per month
Peak concurrent users
—
—
Yearly AWS cost
—
12 months, AWS invoice only

Where the money goes

Monthly AWS cost by category at the current settings. Hover a segment for detail.

Cost curve — how it scales

Monthly AWS total from 1,000 to 100,000 MAU with the current usage assumptions. The dot marks your setting.

Itemized estimate

Every line item at the current settings — the production build-out of the architecture running today.

Pricing pulled 24 Aug 2026 from the AWS Pricing API and published rate cards (us-east-1, USD). AWS list prices change over time — treat this as a snapshot, not a quote, and re-check current pricing before contracting or budgeting from it.

Our recommendation

—

  1. Keep the managed-container architecture for production. It is exactly what the platform runs on today — Fargate containers, RDS PostgreSQL, ElastiCache — hardened for production with Multi-AZ failover, a cache replica, and WAF. AWS operates the moving parts: automated failover, patching, and verified backups, the failure modes that actually hurt a self-hosted platform.
  2. Monitor real traffic before committing to any longer-term pricing. We track actual monthly active users and peak concurrent users on the live environment, so purchasing advice (e.g. reserved/committed pricing) is based on measured demand rather than a forecast. Run on on-demand pricing until that usage picture is clear.

Assumptions & sources

  • Architecture: the managed-container stack the platform runs on today (Fargate + RDS + ElastiCache), production-hardened with Multi-AZ database, cache replica, and WAF. On-demand pricing, one environment (production). Excludes tax.
  • Sizing: peak concurrency modeled as a share of MAU (slider); capacity sized with ~2× headroom over expected peak, autoscaling to 2× steady-state at burst. Capacity constant (~130 concurrent users per vCPU) validated by a load test against the live 1 vCPU / 2 GB API task: it sustained ~55 req/s at ~88% CPU with the database near-idle, i.e. ~165–275 concurrent users per vCPU — the constant sits conservatively inside that band.
  • Key rates (us-east-1): Fargate ARM $0.03238/vCPU-hr + $0.00356/GB-hr; RDS Multi-AZ = 2× single-AZ + gp3 $0.115/GB-mo; CloudFront $0.085/GB first 10 TB then $0.080, 1 TB/mo always-free; MediaConvert ≈ $0.045 per source-minute; SES $0.10/1,000 emails.
  • Widest variable: video egress (CloudFront) — it is the one usage-driven line; everything else is fixed-rate arithmetic.
  • Pulled 24 Aug 2026 from the AWS Pricing API and aws.amazon.com rate cards. Current-environment configuration (sidebar) read live from the running deployment the same day. AWS pricing changes over time — a snapshot, not a quote.