Every request,
one view.

Distributed tracing, real user monitoring and live alerting — owned by your team, never rented.

Inside the Product

One pane of glass for every request, span and pageview.

Distributed tracing, real-user monitoring, infrastructure metrics and AI-driven anomaly detection — unified in a single Novosoft dashboard your team owns.

N
novosoft-api · production Singapore · ap-southeast-1 · Java 21
Overview Transactions Database Traces JVM Service Map
Last 30 min 2
Apdex Score
0.97
SLA met
p99 Latency
42ms
↘ 3.2%
Throughput
1.24k rpm
↗ 8.4%
Error Rate
0.03%
↘ 1.1%
Active Users
2.4k
↗ 12.6%
Uptime · 30d
99.97%
SLO ok
Response time by component
App code PostgreSQL Redis External
Requests by service
15m
  • checkout-svc38%
  • users-svc27%
  • search-svc17%
  • payments-svc14%
  • others4%
Errors by region · 24h
2 alerts
SG
HK
TW
JP
EU
US
BR
247Total errors
92%Recovered
HKTop region
Apdex breakdown
15m
Satisfied91%
Tolerating7%
Frustrated2%
7-day trend
+0.04
Top exceptions · 24h
Live
  • NullPointerException 142
    OrderService.confirm()
  • TimeoutException 98
    PaymentClient.charge()
  • SQLException 54
    UserRepository.findAll()
  • JedisConnectionException 21
    CacheLayer.get()
Slowest transactions
Top 6 avg / p99
EndpointMethodCallsAvgp99ErrorsThroughput
/api/checkout/confirmPOST3,240112 ms412 ms0.04%
/api/payments/chargePOST1,82096 ms284 ms0.18%
/api/searchGET8,54034 ms118 ms0.02%
/api/users/meGET12,10818 ms52 ms0.01%
/api/ordersGET5,42042 ms134 ms0.03%
/api/reports/exportPOST312820 ms2.1 s0.32%
2 active alerts · 18.7M spans / hour · 2,402 sessions live · SLO budget · 78% remaining
Why Novosoft APM

Observability that acts, not just watches.

A full-stack APM your team owns and operates — distributed tracing, real-user monitoring and AI-driven anomaly detection in one self-hosted platform. Datadog-class capability without the per-host bill.

Trace · /api/checkout 412 ms
api-gateway118 ms
order-svc96 ms
payment-svc62 ms
postgres284 ms
redis4 ms

Distributed Tracing

Auto-correlated spans across every microservice. The slow database call hidden three hops down the stack — found in seconds, not hours.

Live Sessions 2.4k online
LCP
1.8s
INP
120ms
CLS
0.12
🇧🇷
São Paulo · Chrome
rage-click on /checkout
Replay

Real User Monitoring

Every real browser and mobile session — Core Web Vitals, journey replay, rage-click detection. Catch UX problems before complaints arrive.

p99 latency · /api/payments AI · auto
Anomaly · 22:34 Likely cause: DB pool exhausted

AI Anomaly Detection

AI-powered outlier detection across metrics, traces and logs — with root-cause hypotheses. Fewer pages, deeper signal, faster fixes.

Own your observability.

Datadog-class APM, RUM and AI-driven anomaly detection — running on your infrastructure, billed once. Your telemetry never leaves your control.

Request a demo