diff --git a/.github/screenshots/readme-carousel.gif b/.github/screenshots/readme-carousel.gif new file mode 100644 index 0000000..da92aa0 Binary files /dev/null and b/.github/screenshots/readme-carousel.gif differ diff --git a/.github/screenshots/ss-01.png b/.github/screenshots/ss-01.png deleted file mode 100644 index 968ad5d..0000000 Binary files a/.github/screenshots/ss-01.png and /dev/null differ diff --git a/.github/screenshots/ss-02.png b/.github/screenshots/ss-02.png deleted file mode 100644 index 84ef201..0000000 Binary files a/.github/screenshots/ss-02.png and /dev/null differ diff --git a/.github/screenshots/ss-03.png b/.github/screenshots/ss-03.png deleted file mode 100644 index 65cc640..0000000 Binary files a/.github/screenshots/ss-03.png and /dev/null differ diff --git a/.github/screenshots/ss-04.png b/.github/screenshots/ss-04.png deleted file mode 100644 index daec35d..0000000 Binary files a/.github/screenshots/ss-04.png and /dev/null differ diff --git a/.github/screenshots/ss-05.png b/.github/screenshots/ss-05.png deleted file mode 100644 index 189894d..0000000 Binary files a/.github/screenshots/ss-05.png and /dev/null differ diff --git a/.github/screenshots/ss-08.png b/.github/screenshots/ss-08.png deleted file mode 100644 index c3f2874..0000000 Binary files a/.github/screenshots/ss-08.png and /dev/null differ diff --git a/.github/screenshots/ss-09.png b/.github/screenshots/ss-09.png deleted file mode 100644 index ab0a634..0000000 Binary files a/.github/screenshots/ss-09.png and /dev/null differ diff --git a/README.md b/README.md index f8479ae..ba8e8ce 100644 --- a/README.md +++ b/README.md @@ -12,29 +12,22 @@ ## Why RunRight -Grafana, Datadog, and Sentry tell you how your application behaves. They do not tell you whether you are paying for the right machine. RunRight answers that one question: given your actual p95 CPU and memory usage, what is the cheapest specific AWS or GCP instance type that still leaves headroom? +Grafana, Datadog, and Sentry tell you how your application behaves. They do not tell you whether you are paying for the right machine. **RunRight solves that: given your actual p95 CPU and memory usage, it recommends the cheapest AWS or GCP instance type that still fits your workload.** -- **No code changes.** It runs as a sidecar process, not an SDK or agent you instrument your app with. -- **Minimal overhead.** Polls `/proc` via gopsutil at a configurable interval (default 5s). Typical footprint: under 5 MB RSS, under 0.1% CPU on the sampler process. Your build time is not affected. -- **Actionable output.** Instead of a graph you have to interpret, you get a specific instance recommendation with a cost delta and a tier label. -- **Works with what you have.** Exports via OTLP, Prometheus, or flat JSON. If you already ship metrics to Datadog or Grafana Cloud, RunRight plugs straight in. +### Key Features + +- **No code changes.** Runs as a sidecar—not an SDK or agent you instrument. Drop it in any CI pipeline. +- **Minimal overhead.** Polls `/proc` every 5s. Footprint: under 5 MB RSS, under 0.1% CPU. Your builds stay fast. +- **Actionable insights.** Get specific instance recommendations with cost delta and tier labels—no graphs to interpret. +- **Works with your stack.** Exports via OTLP, Prometheus, or JSON. Integrates with Datadog, Grafana Cloud, or any observability tool. --- ## In action -
Per-run CPU & memory samples
-
-
-
-
Instance recommendations by tier: right-sized / cheaper-option / more-headroom
- -All jobs · Run history per job
+Live product tour: dashboard, run insights, recommendations, and job history
-
-
+
Pipe your data into Grafana, Datadog, or any metrics tool. Raw runs land in Postgres.