Compute Allocation & Consumption

Edilitics uses dedicated cloud compute to run your Replicate and Transform workflows. Every subscription includes a monthly base allocation, and you can scale elastically with non‑expiring add‑on packs. This guide explains plan characteristics, how compute is consumed, what happens when you run out, and how pricing works by tier.

Plan Characteristics (Machine & Max Runtime)#

Each plan tier maps to a fixed machine profile and a governance cap on maximum runtime per flow. These are enforced for both Replicate and Transform.

Larger machines improve throughput and stability for complex pipelines, but your governance caps (max runtime per flow) still apply.

Base Allocation by Plan#

Base compute resets on the 1st of every month and does not roll over.

Individual (per user)#

• Launch: 15 minutes / month
• Scale: 30 minutes / month
• Pinnacle: 60 minutes / month

Team (pooled across users)#

• Launch: 15 minutes per user / month (pooled)
• Scale: 30 minutes per user / month (pooled)
• Pinnacle: 60 minutes per user / month (pooled)

Pooling: Team allocations combine into a single workspace pool. Consumption is deducted from the pool regardless of who runs the flow.

Free Trial (14 days)#

• Individual Trial: 15 minutes total (non‑pooled).
• Team Trial: 15 minutes total, pooled across up to 2 users.

Add‑ons are not available during the trial period.

Add‑On Compute Packs (Non‑Expiring)#

When the base allocation runs out, buy add‑on packs to continue running workflows. Packs are one‑time purchases that never expire and are consumed after the base monthly allocation.

Scope: On Team plans, add‑on compute is workspace‑wide (shared). On Individual plans, add‑on compute is dedicated to that user.

Launch (2 vCPU • 8 GB)#

Scale (4 vCPU • 16 GB)#

Pinnacle (8 vCPU • 32 GB)#

How Compute is Consumed (Lifecycle)#

Compute minutes are deducted based on actual execution time of a run.

• Setup - Secure VM provisioning, credential decryption, workflow initialization.

• Execution - Extract → Process/Transform → Load (source to destination).

• Teardown - Log final status, securely deallocate compute.

All three stages are counted toward billed compute.

When Compute is Exhausted#

• Active & Scheduled Flows Pause: All Replicate and Transform schedules are automatically paused.

• Notifications: Admins and users receive alerts (in‑app and email; Slack if configured).

• Resume is Manual: After purchasing add‑ons or after the monthly reset, admins must manually resume each paused flow. This prevents accidental overruns.

• Overage Buffer: If a run crosses the remaining balance mid‑execution, a grace buffer up to 15 minutes may be applied. Beyond that, the job is safely stopped.

Consumption Order & Pooling Rules#

• Base first, then add‑ons: Monthly base allocation is consumed first. When it reaches zero, consumption automatically draws from add‑on packs.

• Team pooling: On team plans, all compute (base + add‑ons) is pooled and shared workspace‑wide.

• Individual scope: On individual plans, compute (base + add‑ons) is dedicated to that user.

Tracking & Governance#

• Usage Dashboard (Org Settings): View total allocated, used, and remaining compute; breakdown by workflow type and run status.

• Run History Logs: Per‑run compute usage and outcomes for optimization and compliance.

• Security: Compute instances are isolated per run; credentials decrypt only at runtime; all activity is captured in Audit Logs.

FAQ#

• Do base compute minutes roll over?

  No. Base minutes reset on the 1st of each month.

• Do add‑on packs expire?

  No. Add‑ons never expire and remain available until consumed.

• Are team resources shared?

  Yes. Compute is pooled across all team members.

• What happens mid‑run if we run out of compute?

  A limited grace buffer (up to 15 minutes) can apply; beyond that, the job is stopped and schedules are paused.

• Why do add‑on prices differ by plan?

  Pricing aligns with the plan’s machine size and associated throughput.

• Who gets notified when compute is exhausted?

  Both admins and the initiating users via in‑app and email (plus Slack, if configured).