psychok
v4.2
// Pricing
One number covers monitoring, GitOps, chaos engineering, and the compliance evidence trail — across every node in every cluster you run. The average customer cuts cluster tooling spend by 4.2x within 90 days of migration, and our SLA credits are capped at 99.99% uptime because we don't sell what we can't guarantee.
// 01 — Tiers
Each column is an operational profile, not a SaaS package. Pod‑hour caps, included integrations, and the SLA ceiling are stated up front so a platform engineer can match a tier to a fleet in under thirty seconds.
Tier 01
For teams consolidating 2–3 monitoring tools into one control plane.
Tier 02
For Series B–D SaaS running production Kubernetes at scale (50–500 nodes).
Tier 03
For multi‑region fleets above 500 nodes with regulated workloads.
// 02 — Cost inspection
Same ingest baselines, same metric dimensions, same 30‑day window. We don't smear competitors — we publish the line items so a FinOps lead can verify the 38% and 61% savings claims directly. Numbers are illustrative mid‑market rates published by each vendor in Q4 2024; bring your own ingest profile to your platform assessment for a custom model.
| Line item (per 1M metrics) | PsychoK Platform | Datadog K8s monitoring | New Relic K8s |
|---|---|---|---|
| Container metrics ingest | $0.18 | $0.29 | $0.47 |
| Distributed traces (per 1M spans) | $0.40 | $0.65 | $0.95 |
| Log ingest (per 1M events, indexed 7d) | $0.07 | $0.10 | $0.25 |
| GitOps / ArgoCD pre‑check | Included | Add‑on, $15 / node / mo | Not available |
| Chaos engineering engine | Included | Not available | Add‑on, $25 / node / mo |
| Compliance evidence export | Included | Manual, no native export | Add‑on, $10 / node / mo |
| Pod‑hour rate cap (ceiling) | $0.04 | Per‑host, no cap | Per‑host, no cap |
| Total at equivalent ingest (sample) | $35,712 / mo | $57,600 / mo | $91,800 / mo |
| Δ vs PsychoK | Baseline | +38% | +61% |
Per‑node pricing on the right is at the vendor's published list rate; volume discounts, multi‑year commits, and bundled discounts were excluded to keep the comparison honest. Bring your last 30 days of ingest logs to your platform assessment and we'll re‑run this table against your actual numbers.
// 03 — FAQ
Not sales objections — the technical ones that come up in platform review and FinOps sign‑off.
We meter the maximum concurrent pod count in each 5‑minute window, not the sum of pod‑lifetimes. A HPA that scales from 200 to 800 pods at peak and back down by night is billed on 800 pods for those minutes — never on the cumulative 14,000 pod‑hours. This matches how your cluster actually load‑balances, and it's why the cap sits at $0.04 / pod‑hour rather than something reactive.
Yes. A Platform tier customer with clusters across AWS, GCP, and Azure gets one consolidated monthly invoice broken out by region and by cloud. Enterprise tier customers can tag spend to a cost center, a business unit, or a charge‑back code at the namespace level — useful when the FinOps team is reconciling shared platform cost against product P&Ls.
SLA credits cap at 99.99% uptime. Below that ceiling, credits are issued as a percentage of the affected month's invoice and applied automatically on the next billing cycle — no ticket, no negotiation. We publish the rolling 90‑day uptime window in the status page so customers can audit credits before they hit their invoice.
Nothing breaks. The Platform tier hard‑caps at 500 nodes by design so the SLA and support tier stay predictable. When you cross that line consistently for 30 consecutive days, your account team moves you to Enterprise — same $0.04 pod‑hour rate, negotiable downward at volume, with a named SRE pod and custom chaos recipes. No emergency migration window.
// 04 — Next step
Thirty minutes with a senior engineer. You leave with a cluster spend audit, a tooling consolidation plan, and a baseline MTTR number — even if you never become a customer.
> book --assessment --slot this_weekMedian first response from a senior SRE: 4 minutes 12 seconds. Berlin · San Francisco · Singapore, follow‑the‑sun.