tracewell_

span://hero

See every request.
Catch every regression.

Tracewell follows each request across every service, queue, and database call — so when latency creeps or an error appears, you see exactly which hop caused it and when it started.

Start freeRead the docs →
9.4M spans/min ingestedp99 ingest latency < 40msOpenTelemetry-native
checkout_api.py
01# pip install tracewell
02from tracewell import trace
03 
04trace.init(service="checkout-api")
05 
06@trace.route
07def place_order(req):
08    return db.orders.insert(req)
trace 7f3c…a91 · 4 spanslive

span://product.console

The whole request, on one screen.

Click any slow request and Tracewell reconstructs its full path — every span, every query, every retry — so you stop guessing which service is to blame.

tracewell/ traces / 7f3c-checkout-a91
Trace detail view: a waterfall of stacked spans for a single checkout request, with service names like gateway, checkout-api, payments, and postgres cascading down a millisecond timeline ruler.
latency & SLOs
Latency dashboard: p50, p95, and p99 response-time line chart with an error-rate sparkline below it and KPI tiles for requests per second and error budget.
live logs
Live log stream: rows of structured log lines with INFO, WARN, and ERROR severity badges and timestamps, correlated to the active trace.

span://how-it-works

Three spans from blind to observable.

  1. step 1 / 3

    Instrument

    Drop in the SDK or point your existing OpenTelemetry exporter at Tracewell. One line of config per service, no agents to babysit.

  2. step 2 / 3

    Trace

    Every request gets a trace ID that follows it across services, queues, and databases. Open any one and read its full waterfall.

  3. step 3 / 3

    Alert

    Set latency and error-rate SLOs in plain terms. Tracewell pages you the moment a service drifts — with the offending trace already attached.

Auto-instrumented across your stack

  • PostgreSQL
  • Redis
  • Kafka
  • gRPC
  • AWS
  • Kubernetes
  • NGINX
  • RabbitMQ
  • GraphQL
  • Stripe

span://capabilities

Built for the moment something breaks.

Distributed tracing

Follow one request from the load balancer to the last database write. Spans are stitched automatically across every service and protocol — no manual context passing.

Error tracking

Exceptions are grouped by root cause and linked to the exact trace that threw them. See the stack, the inputs, and every downstream call that ran first.

Latency SLOs

Define p95 and p99 targets per endpoint and watch your error budget burn in real time. Tracewell forecasts the breach before your users feel it.

span://pricing

Priced per span, not per host.

You only pay for the data you actually send. Scale a service to zero and your bill follows.

Free

$0/ forever

For side projects and getting a feel for tracing.

Start free
  • 1M spans / month
  • 3-day trace retention
  • 1 service · 2 seats
  • Community support
Most teams

Team

$90/ month

For teams running real traffic in production.

Start 14-day trial
  • 100M spans / month
  • 30-day trace retention
  • Unlimited services
  • Latency & error SLOs
  • PagerDuty + Slack alerts
  • Priority support

Enterprise

Custom

For org-wide observability with compliance needs.

Talk to us
  • Volume span pricing
  • Custom retention
  • SSO / SAML & audit logs
  • Self-hosted ingest option
  • 99.95% uptime SLA
  • Dedicated engineer

span://get-started

Your next outage already started.
Trace it before it spreads.

Instrument your first service in six lines and watch the spans arrive in seconds. No credit card, no sales call.

Start freeRead the docs →