Every Deployment. Every Rollback. Every Change — On One Timeline.
Stop piecing together deploy history from CI logs, Slack threads, and memory. OpsTrails is deployment software that records changes instead of running them — every deploy, rollback, release, and data load, queryable by your team or your AI assistant. Not logs. Not metrics. The changes you chose to make.
1,000 change events/month free — that's ~30 deploys a day, org-wide. 5 minutes to integrate, one API call per pipeline. See how it works ↓
Plugs Into Your Existing Pipeline in 5 Minutes
One API call per pipeline. No agents to install, no infrastructure changes. OpsTrails works with whatever you already use.
GitHub Actions
Add one step to your workflow YAML.
GitLab CI
One job in your .gitlab-ci.yml.
Bitbucket Pipelines
One pipe in your bitbucket-pipelines.yml.
Jenkins
One POST in your Jenkinsfile.
HTTP API
Works with any tool that can make an HTTP call.
TypeScript SDK
Type-safe client for Node.js and Deno.
CLI
Record events from any script or terminal.
Seconds, not archaeology
One timeline query replaces piecing together deploy history from CI logs, Slack threads, and memory during an incident.
No new dashboard to learn
Engineers ask their AI assistant “what changed?” via MCP and get the answer in natural language.
Running in minutes
One API call per pipeline. No agents to install, no infrastructure changes.
Your Deployment History, Queryable and Complete
Every deployment, rollback, release, and data load — with timestamps, sources, versions, and environment context.
DEPLOYMENT TIMELINE — TODAY
09:15 — api-gateway deployed v3.2.0 to staging
Source: GitHub Actions
10:30 — api-gateway deployed v3.2.0 to production
Source: GitHub Actions · Promoted from staging
11:45 — user-service deployed v1.8.4 to production
Source: GitLab CI
14:32 — payment-service deployed v2.4.1 to production
Source: GitHub Actions
14:38 — payment-service error rate spike — before: 0.12%, after: 2.41%
Metric: Sentry error_rate · Auto-correlated
14:41 — payment-service rolled back to v2.4.0
Error rate returned to 0.14% · Incident duration: 9 min
Every event searchable by service, environment, time window, or version. Query with AI via MCP — or use the dashboard and API.
An Event Is a Decision, Not a Log Line
OpsTrails isn't a log feed. Your observability stack ingests millions of data points about what your system does. OpsTrails records the far smaller set of moments when someone changes it — a deploy goes out, a rollback fires, a data load runs, an index rebuilds.
Log & metric feeds
Millions of machine-generated data points per day. Volume grows with traffic. Sampled, noisy, priced by ingestion.
OpsTrails change events
Dozens of deliberate changes per day — even org-wide. Volume grows only with how often you decide to change something. Complete, human-scale, AI-queryable.
The math at scale: a 200-service organization shipping 50 changes a day records about 1,500 events a month — not millions. Event volume tracks decisions, not data, so the timeline stays readable and affordable no matter how big your infrastructure gets.
The More Services You Run, the More the Timeline Matters
With 5 services you can hold the day's changes in your head. With 50 or 200 — spread across teams, pipelines, and environments — you can't. That's exactly when a single change timeline becomes essential.
One timeline, every team
Every pipeline gets its own API key, and every team writes to the same timeline. GitHub Actions here, Jenkins there — it all lands in one queryable history.
Cross-service causality
user-service breaks twenty minutes after api-gateway deploys. When changes from every service sit on one timeline, that connection is a query — not a war room.
Governance built in
Deployment policies and change-freeze windows put guardrails around when changes may ship — and the timeline doubles as your audit trail of what actually happened.
Track the Metrics That Matter — Automatically
OpsTrails calculates DORA metrics from your deployment data. No manual tagging, no spreadsheets, no custom dashboards.
Deployment Frequency
How often your team ships to production. Track trends across services and teams. Elite teams deploy on demand, multiple times per day.
Change Failure Rate
What percentage of deployments cause incidents or require rollbacks. Elite teams achieve 0–15%, low performers sit at 45–60%.
Mean Time to Recovery
How fast you recover when a deployment goes wrong. With OpsTrails, root cause identification drops from hours to seconds.
Lead Time for Changes
From commit to production. Understand where bottlenecks hide in your delivery pipeline and measure improvement over time.
Know where your team stands. Measure improvement over time. Read about DORA metrics →
See the Impact of Every Deployment
Connect your monitoring tools and OpsTrails automatically compares metrics before and after each deployment. See exactly how every release affected error rates, latency, and availability — without manual correlation.
See how OpsTrails works alongside Datadog, New Relic, or PagerDuty.
Ask Your AI Assistant, Not Your Team
OpsTrails exposes your deployment timeline through the Model Context Protocol (MCP). Any compatible AI assistant can query your deployment history in natural language.
$ “What was deployed to production today?”
3 deployments today: api-gateway v3.2.0 at 10:30, user-service v1.8.4 at 11:45, and payment-service v2.4.1 at 14:32 (rolled back to v2.4.0 at 14:41 due to error rate spike).
$ “Which service was last rolled back and why?”
payment-service was rolled back from v2.4.1 to v2.4.0 at 14:41 today. Sentry error_rate went from 0.12% to 2.41% after the deployment. Error rate returned to 0.14% after rollback.
$ “Show me all deployments with error rate increases this week”
Found 1 deployment with error rate increase: payment-service v2.4.1 on Friday — error_rate went from 0.12% to 2.41% (+1,908%). This deployment was rolled back within 9 minutes.
Frequently asked questions
Isn't OpsTrails another logging tool?
No. OpsTrails doesn't collect logs, traces, or metrics. It tracks operational events — deployments, rollbacks, data loads, and releases — on a single timeline. Think of it as a changelog for your infrastructure that your AI assistant can query.
How is OpsTrails different from Datadog?
Datadog monitors application performance. OpsTrails tracks what changed and when. They're complementary — connect Datadog as an analytics provider and OpsTrails automatically shows before/after metrics for every deployment.
Does OpsTrails replace monitoring tools?
No. OpsTrails sits alongside your monitoring stack. It answers 'what changed?' while your monitoring answers 'what's broken?' Connect them together for automatic impact analysis.
How long does setup take?
Under 5 minutes. Add one step to your CI/CD pipeline to start recording events. Connect an analytics provider for impact metrics. Set up the MCP server for AI queries. Each step is independent.
What happens when I exceed my event limit?
New events are rejected with a 429 response until the quota resets on the 1st of the month. You'll receive a notification at 80% usage so you can upgrade proactively. Events already recorded stay accessible for the full retention period.
Related Reading
Start Tracking Every Deployment Today
Free tier includes 1,000 change events/month — enough to evaluate across a real fleet — plus 100 MCP calls and built-in DORA metrics. No credit card required. Paid plans scale to 50 teammates, 50 pipelines, deployment policies, and change-freeze windows.