Infrastructure That Scales
With Your Organization
From a single server pilot to a multi-region Kubernetes cluster. Right-size your deployment, understand your cloud costs, and grow without re-architecting.
Sizing Calculator
Enter your member count and concurrent users to get a tier recommendation and resource estimate.
Modeled estimate — pilot deployments validate assumptions.
Hardware Specifications by Tier
Detailed infrastructure requirements for each deployment tier.
| Specification | Starter 10K – 50K members | Professional 50K – 250K members | Enterprise 250K+ members |
|---|---|---|---|
| Deployment | Single-node Docker Compose | Clustered (3–5 nodes) | Kubernetes (5–10+ worker nodes) |
| Nodes | 1 server | 3–5 servers | 5–10+ worker nodes |
| CPU / Node | 4–8 CPU cores | 8 CPU cores | 8–16 CPU cores |
| RAM / Node | 16–32 GB | 32 GB | 32–64 GB |
| Total CPU | 4–8 cores | 24–40 cores | 40–160+ cores |
| Total RAM | 16–32 GB | 96–160 GB | 160–640+ GB |
| Storage | 500 GB NVMe SSD | 1 TB NVMe SSD per node | 5+ TB distributed (SSD) |
| Services | ~25 | ~35 | ~51 |
| Setup Time | 2–3 weeks | 4–8 weeks | 8–12 weeks |
| Database & Cache | |||
| PostgreSQL | PostgreSQL 16 (single instance) | PostgreSQL 16 (primary + replica) | PostgreSQL 16 HA (Patroni / CloudNativePG) |
| Max Connections | 100 | 300 | 500 |
| Replication | None (daily backups) | Streaming replication | Synchronous multi-replica + Velero backup |
| Redis | Standalone (384 MB) | 3-node Sentinel cluster (1 GB) | 6-node Redis Cluster (3 primary + 3 replica) (4+ GB) |
| Kafka | Apache Kafka (single broker) | Apache Kafka (3-broker cluster) | Apache Kafka (5+ broker cluster / Strimzi) |
| Reliability | |||
| High Availability | — | ||
| Auto-Scaling | — | — | |
| Availability SLO | 99.5% | 99.9% | 99.99% |
| P95 Latency | <300 ms | <200 ms | <150 ms |
| Throughput | 200 req/s | 1,000 req/s | 5,000+ req/s |
* All capacity numbers are modeled estimates. Pilot deployments validate assumptions for your workload.
Cloud Infrastructure Costs
Estimated monthly costs for compute, database, cache, and networking.
Infrastructure costs only — excludes HDIM license fees. See Pricing for license details.
Starter
Professional
Enterprise
Estimates based on on-demand pricing as of March 2026. Reserved instances or committed-use discounts can reduce costs 30–50%. Data transfer and storage egress fees not included.
Performance Benchmarks
Real load-test results from HDIM deployments.
Single Instance
- Cached throughput
- 400–500 req/s
- Uncached throughput
- 150–200 req/s
- P95 (cached)
- 180 ms
- Concurrent users
- 100
3-Node Cluster
- Sustained (300 users)
- 900 req/s
- P95 sustained
- 240 ms
- Peak (600 users)
- 1,500 req/s
- CPU at sustained
- 65%
Cache Impact
- Cold (0% hit)
- 150 req/s req/s
- Warm (50% hit)
- 280 req/s req/s
- Hot (90% hit)
- 450 req/s req/s
- Warmup time
- 15 min
HEDIS Measure Latency (Cached)
Growth Path
Start small, scale with confidence. Each tier builds on the previous one.
Starter
Prove ROI with a pilot
- Members
- 10K – 50K
- Services
- ~25
- Infrastructure
- 1 server
- Availability
- 99.5%
- Setup
- 2–3 weeks
Professional
Production-grade with HA
- Members
- 50K – 250K
- Services
- ~35
- Infrastructure
- 3–5 servers
- Availability
- 99.9%
- Setup
- 4–8 weeks
Enterprise
Unlimited scale with K8s
- Members
- 250K+
- Services
- ~51
- Infrastructure
- 5–10+ worker nodes
- Availability
- 99.99%
- Setup
- 8–12 weeks
Migration Procedures
Starter → Professional
Low Risk- 1.Provision 2–4 additional servers
- 2.Set up PostgreSQL streaming replication
- 3.Deploy Redis Sentinel (3 nodes)
- 4.Configure load balancer (HAProxy / NGINX)
- 5.Migrate Docker Compose to multi-node with Swarm or Ansible
- 6.Enable additional service profiles (ai, analytics)
- 7.Validate HA failover with chaos testing
Professional → Enterprise
Medium Risk- 1.Provision Kubernetes cluster (EKS / GKE / AKS)
- 2.Convert Docker Compose to Helm charts
- 3.Migrate PostgreSQL to managed HA (Aurora / AlloyDB / Azure DB)
- 4.Deploy Redis Cluster (6 nodes) or managed equivalent
- 5.Set up Kafka on Kubernetes (Strimzi) or managed service
- 6.Configure HorizontalPodAutoscaler for all services
- 7.Implement NetworkPolicies and PodDisruptionBudgets
- 8.Set up Velero for disaster recovery
- 9.Run full load test at target capacity
Ready to Right-Size Your Deployment?
Our engineering team will help you design the optimal infrastructure for your member population, compliance requirements, and growth plans.