Know whether a release is safe. Know what broke. Know the safest way back.
MOJAQ Change Guard evaluates code, flag, model and configuration changes before release, watches the rollout, and correlates symptoms into evidence-backed hypotheses when something regresses. It can recommend reversible next steps, while people retain approval and uncertain evidence stays explicitly uncertain.
See it catch a bug. In real time.
A deploy ships, something breaks, and MOJAQ correlates the deploy, errors, logs, uptime and an investigation hypothesis in one view — then take the tour through Rewind, Tracing and Compliance. The live counter is genuinely from our realtime service.
A guided tour of four tools — MOJAQ ships 20+. See them all →
Error rate
Live logs · payment-service
What's happening
/checkoutAnalyzing…
MOJAQ joins deploys, errors, uptime and flags automatically.
Error rate 21× baseline after deploy v2.4.0
Release intelligence needs evidence, not another chat window
Analytics, uptime, errors, logs, traces, real-user signals, flags, deploys and AI usage feed a shared correlation layer. They remain useful tools, but their primary job is to help Change Guard compare a release with its baseline and explain what the available evidence does and does not support.
Analytics
Cookieless aggregate pageviews and custom events. Disclosure and consent requirements still depend on your use case.
Open analytics →Uptime
HTTP & port monitors with response-time graphs, incident timelines and public status pages.
Open monitors →Error Tracking
Sentry-compatible DSN with smart grouping, AI-assisted hypotheses and deploy correlation. Keep your SDKs.
Open errors →Render
Turn any URL or HTML into a pixel-perfect PDF or PNG with one API call, invoices, reports, screenshots.
Open render →Feature Flags
Boolean flags and percentage rollouts with targeting, toggled instantly without a deploy.
Open flags →Realtime separate
Pusher-protocol WebSockets, public, private and presence channels with broadcasts and a debug console.
Open realtime →Tracing
Distributed traces with span waterfalls and p50/p95 latency. Send one JSON batch, see exactly where the time goes.
Open tracing →AI Gateway new
OpenAI-compatible gateway with budget caps, kill-switch and automatic model fallback. Point your client, keep your code.
Open AI gateway →Rewind new
Ranked hypotheses and reversible next steps, correlated across deploys, flags, AI calls and real users. Evidence first, causal claims only when supported.
See the demo →Domain Watch
Every SSL cert and subdomain from Certificate Transparency + DNS. Catch expiries and subdomain-takeover risk before attackers do.
Open domain watch →Compliance
Create DPIA, AI Act and data-flow working drafts from limited observed usage. Unknowns and required review remain explicit.
Open compliance →Real-User
Core Web Vitals and JS errors from real browsers, cookieless. LCP, INP and CLS at p75, broken down per page.
Open real-user →From release question to defensible answer
MOJAQ's services use several purpose-built databases behind a shared correlation layer. That layer connects changes, symptoms and evidence without pretending the architecture is a single datastore.
It shows what changed and what symptoms followed
"Errors spiked 4 minutes after deploy v2.4.0 while flag new-checkout was on." Sentry can't see your deploys, UptimeRobot can't see your flags. MOJAQ correlates the available records without treating timing as proof of cause. Watch it above ↑
Ask Claude & Cursor about your app
The MOJAQ MCP server lets your AI assistant query your deploys, errors, logs and uptime, read-only, right in your editor. See the MCP server →
One API key and one evidence trail
Connect operational signals under one account and review the disclosed processor paths and DPA working template before relying on contractual claims.
Status pages & README badges included
Publish a hosted status page and a live uptime badge in a click, the kind other vendors charge $20+/mo for. See an example →
A Helsinki-hosted alternative to self-hostable observability
Many teams look at self-hostable observability because data location and processor sprawl matter. MOJAQ is a managed alternative with core service storage in Finland and documented international paths, without requiring you to operate storage, upgrades and the correlation layer.
- EU-first architecture. Core storage location and international processing paths are documented on the Trust page.
- DPA working template. Publicly reviewable, with fields and safeguards that must be completed before execution.
- One platform, not a vendor braid. Every tool under one account, one dashboard, one REST API.
If you can call a REST endpoint, you can use MOJAQ
Drop one script tag in for analytics. Call render from any language. Ship a deploy marker from CI. Every tool speaks the same documented API, authenticated with one key. Pick any feature to see the exact snippet, most take about two minutes.
- One X-API-Key across every product
- Sentry & Pusher protocol compatible, keep your SDKs
- Copy-paste examples in the dashboard for every endpoint
One free package while we're in beta
During the beta there's exactly one package, Beta Access, and it's free. Every tool, generous limits, no credit card, no commitment.
Paid plans will be announced in advance, with plenty of notice, never a surprise bill.
Everything in the suite, one account.
- Analytics, Uptime, Errors, Render & Flags
- Pusher-compatible Realtime included
- Full REST API & one shared key
- Core EU storage, subprocessors disclosed, DPA working template
No card required · cancel anytime · your data is yours
Questions teams ask before sending us production data
Where is MOJAQ hosted?
Is MOJAQ free during the beta?
Which developer tools are included?
Does MOJAQ support GDPR obligations?
Can I keep my existing Sentry or Pusher clients?
Start watching your stack in two minutes
Create an account, grab your API key, drop in one script tag. That's it.