About
Greenslope is a UK-registered, founder-run company building release-aware observability with a closed incident loop — for small engineering teams shipping on PaaS. This page explains who we are, what we believe, and the shape of the commitments we make.
Observability should close the loop, not just observe.
Roughly two-thirds of software outages are caused by configuration or change (Uptime Institute, 2023). General-purpose observability treats releases as an afterthought and assumes the customer has an SRE org to drive triage. Most small teams don't.
Three shifts make Auto-SRE possible now. OpenTelemetry has reached vendor-neutral maturity, so ingest is a solved problem rather than a per-language engineering project. LLM inference has commoditised to the point that automatic triage runs at under a penny per alert. And AI-assisted coding has compounded the shipping-velocity problem — small teams ship faster than ever while SRE capacity has not grown.
Greenslope is the consequence. Detect the incident, triage it, suggest the rollback, draft the postmortem — then tune from how the team responded. The engineer on-call gets answers, not a canvas. The team gets reliability practice without having to invent it.
Our 24-month category target is that “Auto-SRE” becomes a recognised category, not a feature bolted onto observability. You’ll know it’s working when evaluators arrive looking specifically for Auto-SRE.
Six principles the product is built around.
Close the loop, don't just observe.
Observability tools stop at detection. Greenslope treats detection as the start of the workflow — triage, rollback suggestion, postmortem. Seeing the problem is necessary; resolving it is the product.
Opinions belong in the product.
Most small teams don't want to become SREs. We propose SLOs, thresholds, and rollbacks so the team doesn't have to invent reliability practice from scratch. You always confirm.
Honest non-goals over feature-matrix arms races.
We list what we don't do — logs, custom metrics, RUM, synthetics, rotations at V1 — alongside what we do. Evaluators self-select better when the shape is clear.
OTel-only, no bespoke SDK.
We ingest standard OpenTelemetry. No Greenslope SDK to learn, no vendor lock-in, no per-language version matrix for a two-person team to maintain.
EU residency by default, not by upgrade.
UK-registered company, single EU processing region, UK↔EU adequacy. No SCCs needed for EU customers. Residency is a day-one promise, not an enterprise add-on.
Human-in-loop, always.
Greenslope suggests rollbacks and surfaces decisions. It does not roll back on its own and does not autonomously change thresholds against your will.
How we write — across product, docs, support, and here.
The alert telling you to roll back uses the same tone as this page. Five tenets, applied consistently.
Direct
"We suggest increasing the latency threshold to 450ms" beats "Based on observed distribution, you may wish to consider…".
Honest
Constraints and non-goals are surfaced, not hidden. Our security page lists what we don't yet have. Our DPA is clearly caveated as a V1 template.
Slightly opinionated
We have views. Auto-SRE over configure-yourself. Release-first triage. We say so. Confident, not aggressive — competitors are peers.
Not jargon-heavy
We assume baseline engineering literacy, not SRE fluency. The first time you see burn-rate or error budget in the product, it comes with a one-line explanation.
Never sarcastic, never apologetic
Sarcasm ages badly. Apology implies a mistake. The tier system is a designed choice — we say so.
Small team. London-based. EU-hosted.
Residency
EU — europe-west1
All customer telemetry processed in a single EU region. UK↔EU adequacy means no SCCs needed for EU customers.
Funding
Bootstrapped
No outside capital at V1. Paddle as Merchant of Record handles billing, tax, and chargebacks worldwide.
Next step
If you want to see whether Greenslope matches your reliability shape, try it free for 14 days — or send a question to a real mailbox a real person answers.