Trust & reliability
Built like infrastructure, because for your company it is
When 10,000 caregivers depend on the platform on a given morning, thirty minutes of downtime is missed care and millions in delayed billings. Here is how Arcapse is engineered — and contractually guaranteed — to not let that happen.
99.99%
uptime SLA on Enterprise, backed by financial credits
15 min
Severity-1 response, 24/7 — a human, not a ticket queue
0
EVV events dropped at the 9:00 AM clock-in spike
14–30 d
to migrate from your old platform, dual-run included
One platform, three isolation tiers
Every request enters through a global API gateway — authentication, tenant routing, rate limits — and lands on the isolation tier that fits the company's size and risk posture. Moving tiers is a migration we run, not a re-implementation you buy.
Global API gateway
Auth · dynamic tenant routing · rate limiting
Pooled multi-tenant
Small & mid companies
- Shared app compute
- Postgres row-level security, forced on every tenant table
- Low cost · onboard in days
Dedicated database tenant
Large companies
- Shared app compute
- Isolated database instance
- High throughput, no noisy neighbors
Isolated VPC tenant
Enterprise networks (5,000+ caregivers)
- Dedicated compute
- Dedicated database in its own VPC
- Full data isolation for corporate risk policies
Available at 8:59 AM, when everyone clocks in at once
Multi-region by design
Canadian tenants run in ca-central-1 with cross-AZ failover; US tenants run active-passive across us-east-1 and us-west-2. PHI never leaves its country — failover stays inside the data region.
Event-driven EVV ingestion
Tens of thousands of caregivers clock in inside the same two minutes. Check-ins land on a durable event stream and are drained by auto-scaling workers — a spike queues for seconds; it never drops a visit.
Offline-first in the field
The mobile app writes to on-device storage before any network call — notes, signatures and clock-ins work in an elevator or a rural dead zone, then sync silently when service returns. Retries never discard local data.
Ready for the 500-question security review
Federated SSO & SCIM
SAML 2.0 with Okta, Microsoft Entra ID or Google Workspace. Staff are provisioned when HR hires and de-provisioned the minute they're terminated — sessions revoked in seconds, never a hand-typed invite.
Custom roles, field-level visibility
Admins build roles like "Intake Coordinator — Region A" on the same permissions matrix the API enforces. Field rules are serializer allowlists: a role that can't see rates never receives them.
Tamper-proof SIEM streaming
Every PHI read, edit, export and delete lands in an append-only ledger the app role cannot update or delete — and streams to your Datadog, Splunk or CloudTrail Lake in near real time.
Your data, your warehouse
An enterprise add-on mirrors live operational data into your Snowflake, BigQuery or Redshift every 15 minutes. Raw tables for your analysts — not PDFs.
TLS 1.2+ in transit · AES-256 at rest · PHIPA, PIPEDA and HIPAA-aligned · SOC 2 program with controls tracked from day one.
Enterprise service, in the contract
Implementation lead
A named engineer runs your migration: automated extraction from the incumbent's exports, custom ETL maps, a sandbox parallel run and a reconciliation report — live in 14–30 days, not months.
Dedicated CSM
Quarterly business reviews against your own metrics, workflow tuning, and a roadmap channel that actually answers. Enterprise accounts never queue behind a chat widget.
Enterprise SLA support
15-minute Severity-1 response around the clock, a public status page with incident comms, and financial credits on your bill if we miss the SLA — in the contract, not the brochure.