Home
FAQ

Sazabi FAQ

By Sazabi Team

Last verified on Nov 8, 2026.

Short answers to the questions people ask about Sazabi, each drawn from the pricing page or the Sazabi docs. If a price or limit here ever differs from the pricing page, the pricing page is right.

Where to Read More

Frequently Asked Questions

What Sazabi Is

What is Sazabi?

Sazabi is AI-native observability and security for engineering teams that ship quickly. It connects your production logs with your code and deployment history. You can ask its agent a question in plain language and see the records and queries behind the answer, or let Sazabi open an issue with supporting evidence when it finds a problem.

How do I use Sazabi day to day?

You send your logs, then either ask or get told. To ask, open a thread and describe the problem; the agent runs the queries, reads the results, and replies with the log lines and correlations it found. To get told, let Sazabi open issues and notify your team. The web dashboard, Slack, Microsoft Teams, your coding agent, the CLI, and the API all reach the same threads and data.

What is an issue in Sazabi?

An issue is a problem the Sazabi agent found in your data and decided is worth your attention. Unlike an alert, it arrives already investigated, with what happened, why, and the evidence. Each issue has a name, a description, a severity, an optional component, and a status of open, resolved, or ignored. There is no monitor behind an issue, so there is no rule set for you to write or keep in sync.

Does Sazabi cover security as well as reliability?

Yes. Security review scans the code in your connected repositories for vulnerabilities and files each confirmed finding as an issue, so it reaches the same inbox and notification rules as runtime issues. It is on by default for every project, and an owner or admin can turn it off per project. Scanning is not billed, but each finding counts toward issue usage like any other issue.

Setup and Data Sources

How long does it take to get started?

The quickstart takes about 10 minutes and ends with a project receiving logs and an answer from the agent grounded in a log you sent yourself. Every step works from the dashboard or the sazabi CLI, and you choose the region where the project stores its data when you create it.

How do I send logs to Sazabi?

There are two setup modes. Native setup connects a supported platform account, such as Vercel, Cloudflare, Google Cloud, or AWS CloudWatch, and Sazabi configures forwarding for you. Manual setup gives you a keyed intake URL that you point a collector, SDK, platform, or application at. You can also instrument backend code with OpenTelemetry.

Does Sazabi take metrics and traces, or only logs?

Sazabi ingests OpenTelemetry metrics (Prometheus metrics through an OpenTelemetry Collector) and OTLP traces through the same intake as your logs. It stores each metric data point and each span as a log record that its agent queries alongside your logs, so there is no separate metrics store to run.

Does Sazabi offer real user monitoring or synthetic checks?

Sazabi's Browser and React Native SDKs capture errors, console output, network calls with trace context, user interactions, navigation and session identity from web and mobile apps. They do not capture Core Web Vitals or session replay. Sazabi does not offer synthetic checks, so keep your synthetic monitoring tool for those.

Which tools does Sazabi integrate with?

The integrations directory lists more than 100 entries. They cover log sources such as AWS, Google Cloud, Vercel, Cloudflare, and OpenTelemetry, chat in Slack and Microsoft Teams, code in GitHub and Bitbucket, and incident tools such as PagerDuty, VictorOps, Rootly, and incident.io. Jira, Linear, PagerDuty, VictorOps, and incident.io need the Startup plan or above.

Can I try Sazabi without replacing my current tools?

Yes. If you already run an OpenTelemetry Collector, Grafana Alloy, Fluent Bit, Vector, or a Datadog Agent, add your project's intake URL as one more destination. Your existing pipeline keeps sending to your current tool, and Sazabi receives a copy.

Pricing and Billing

How much does Sazabi cost?

Plans are priced per 30-day cycle: Hobby is $50, Startup is $200, and Growth is $1,000, and Enterprise pricing is custom. Hobby, Startup, and Growth include the same usage allowances and overage rates; they differ in workspace limits, log retention, features, and support. There is no yearly billing or annual discount.

Is there a free plan?

No. Every organization is on a paid plan, and Hobby at $50 per 30-day cycle is the smallest. Canceling takes effect at the end of your current billing cycle, and service continues until then.

What is included in each plan, and what does overage cost?

Hobby, Startup, and Growth each include 200 GB of logs ingested, 50 GB of logs scanned by queries, 20 issues, 500 AI credits, and 5,000 automation runs per cycle. Past that, ingested and scanned logs cost $0.10 per GB, issues cost $4 each, AI credits cost $0.01 each, and automation runs cost $0.01 each.

How do AI credits work?

Credits pay for the tokens the Sazabi agent uses as it investigates. Your plan adds credits each cycle, and unused credits carry forward and do not expire. You can buy more with a one-time or automatic top-up, and a cycle spending limit caps automatic top-ups. If your balance runs out, issue detection and log ingestion continue, but new AI investigations wait until you add credits.

Do I pay for an issue Sazabi got wrong?

Not if you reject it in time. You can reject an issue within 24 hours of its creation, and Hobby, Startup, and Growth each include 20 rejections per cycle. A rejected issue is excluded from billable issue usage and cannot be restored. Ignoring an issue keeps it, and keeps it billable.

Security, Data, and Compliance

Does Sazabi train models on my data?

No. Sazabi is a data processor and works with your data only to run the product for you. Your data is never sold and never used to train models, either Sazabi's or its providers'. Sazabi uses enterprise API tiers that contractually exclude submitted content from training and passes that commitment to you in its Data Processing Addendum.

Where is my data stored, and for how long?

You choose a United States or European Union region when you create a project, and it is fixed for the life of the project. Data is encrypted with TLS 1.2 or higher in transit and AES-256 at rest. Logs are kept for your plan's retention window, 7 days on Hobby, 30 on Startup, and 90 on Growth, then deleted automatically. Enterprise retention is custom.

Can Sazabi redact sensitive data from my logs?

Yes. Turn on Redact sensitive data in project settings and Sazabi replaces email addresses, phone numbers, US Social Security numbers, payment card numbers, and secrets with placeholders before logs are stored or shown to the agent. It is off by default, applies only to logs received after you turn it on, and does not catch free-form personal data such as names or addresses. For data that must never reach Sazabi, strip it at your log source.

Which certifications does Sazabi hold?

Sazabi holds SOC 2 Type II, SOC 1, ISO 27001 and ISO 27017 certifications. Anyone who signs an NDA, customer or not, can request the SOC 2 Type II report and ISO 27001 audit reports from the Sazabi Trust Center (sazabi.secureframetrust.com). HIPAA safeguards are available on select plans. Sazabi also supports GDPR and CCPA, signs a Data Processing Addendum, and has independent third parties run annual penetration tests.

Can I run Sazabi in my own cloud?

Yes, with bring your own cloud (BYOC). Log intake, processing, and storage run in your own AWS account, and your logs live in a database account your organization owns, while Sazabi deploys, upgrades, monitors, and supports it. Sazabi staff have no standing access, and you can revoke access at any time. BYOC is available under an enterprise agreement.

Does Sazabi support single sign-on?

Yes. Sazabi supports SAML 2.0 and OpenID Connect with identity providers such as Okta, Microsoft Entra ID, and Google Workspace, plus SCIM provisioning so your identity provider can add, update, and remove members. Single sign-on needs a plan that includes it, so contact Sazabi to turn it on.

Coding Agents and MCP

Can my coding agent use Sazabi?

Yes. Once connected, your agent can search your production logs, look up issues, and run the same investigations you would run in a thread. Terminal agents such as Claude Code, Codex, and OpenCode use the sazabi CLI. GUI clients such as Claude Desktop and ChatGPT connect through the MCP server.

What is the Sazabi MCP server?

It is a Streamable HTTP server at https://mcp.sazabi.com/mcp that exposes the Sazabi API through two tools, search and execute, so new API operations become reachable without new tools. Authenticate with a secret key in the X-Sazabi-Secret-Key header, or with OAuth in clients that support it. Secret keys grant read and write access, so keep them out of source control.

Comparisons

How does Sazabi compare with Datadog, New Relic, and other tools?

Sazabi publishes a comparison for each major tool, with list prices and facts checked on each vendor's own pages and a dated source list. They cover how each product meters data, retention, AI investigation, and deployment options, and when the other tool is the better fit. Start from the comparison hub at sazabi.com/compare.

Sources

  1. What Sazabi is, ask or get told, answers with the queries the agent ran, the surfaces that reach the same threads (accessed Nov 8, 2026)
  2. Everything is a log: metrics, traces, and webhook events normalize into logs through one intake (accessed Nov 8, 2026)
  3. Issue definition, issues compared with alerts and monitors, issue fields and statuses (accessed Nov 8, 2026)
  4. Security review scans connected repositories, on by default, findings filed as issues, scanning not billed (accessed Nov 8, 2026)
  5. About 10 minutes to a project receiving logs and a first answer, dashboard or CLI for each step, region choice (accessed Nov 8, 2026)
  6. Native and manual setup modes, connecting a supported platform, adding the intake URL to an existing collector or Datadog Agent (accessed Nov 8, 2026)
  7. The integration directory (120 entries on Nov 8, 2026), Slack, Microsoft Teams, GitHub, PagerDuty, VictorOps, Rootly, and incident.io (accessed Nov 8, 2026)
  8. Plan prices per 30-day cycle, included usage, overage rates, retention per plan, issue rejections, plan features, no yearly billing (accessed Nov 8, 2026)
  9. Every organization is on a paid plan, issue rejection within 24 hours, rejection excludes the issue from billable usage and cannot be undone, cancellation at cycle end (accessed Nov 8, 2026)
  10. Credits cover agent token usage, unused credits do not expire, one-time and automatic top-ups, cycle spending limit, issue detection continues at zero credits (accessed Nov 8, 2026)
  11. No model training on customer data, enterprise API tiers, TLS 1.2 or higher and AES-256, US or EU region, deletion, redaction at intake and its limits (accessed Nov 8, 2026)
  12. Managed storage, logs, metrics, traces, and events stored as searchable records, retention applies going forward (accessed Nov 8, 2026)
  13. GDPR and CCPA support, Data Processing Addendum, subprocessor list, annual third-party penetration tests, security@sazabi.ai (accessed Nov 8, 2026)
  14. BYOC: log intake, processing, and storage in your AWS account, what stays with Sazabi, no standing staff access, revocable access (accessed Nov 8, 2026)
  15. SAML 2.0 and OpenID Connect, Okta, Microsoft Entra ID, and Google Workspace, SCIM provisioning, plan requirement (accessed Nov 8, 2026)
  16. CLI path for terminal agents (Claude Code, Codex, OpenCode), MCP path for GUI clients (Claude Desktop, ChatGPT), secret keys (accessed Nov 8, 2026)
  17. MCP server URL, X-Sazabi-Secret-Key header, OAuth support, the search and execute tools (accessed Nov 8, 2026)
Send your first logs to Sazabi