SaaS application monitoring on autopilot

Automatic anomaly detection, endpoint health, end-to-end tracing, error monitoring and alerting, and background job monitoring out of the box. Connect your coding agent through MCP to investigate.

Node.js Python OpenTelemetry

Features

Everything you need to watch production.

~/payments-api/errors/eg_a41f

ResolveMute

UpstreamTimeoutErrorexceptioneg_a41f

connect ETIMEDOUT acquirer-gw.internal:443

Error: connect ETIMEDOUT acquirer-gw.internal:443 after 5000 ms

EndpointPOST /v1/payments/:id/capture

First seen38 minutes ago

Last seen2 minutes ago

Occurrences52

Sampled traces16

Affected endpoints1

Occurrence trend

52 occurrences · peak 10/bucket

occurrences

Stack trace

Error: connect ETIMEDOUT acquirer-gw.internal:443 after 5000 ms
  at TCPConnectWrap.afterConnect (node:net:1595:16)
  at AcquirerClient.capture (src/lib/acquirer.ts:87:11)
  at capturePayment (src/routes/payments/capture.ts:42:20)
  at processTicksAndRejections (node:internal…)
  at async POST /v1/payments/:id/capture (src/router.ts:118:9)

~/payments-api/traces/7a3f9c1e

prodlive

ERRORtrace 7a3f9c1e

POST /v1/payments/:id/capture

Duration5.21 s

Spans6

Started2 m ago

Waterfall

0 ms1.30 s2.61 s3.91 s5.21 s

serverPOST /v1/payments/:id/capture

5.21 s

internalmiddleware.auth

12 ms

clientdb.query SELECT payments

31 ms

internalAcquirerClient.capture

5.00 s

clienttcp.connect acquirer-gw.internal:443SLOW

5.00 s

clientpostgres.query UPDATE payments

9 ms

payments-api·client

postgres.query

statuserrorduration9 msoffset+5.13 sspan idc4e19b02

Attributes5Events1Related1

Database

db.systempostgresql

db.statementUPDATE payments SET status=$1 WHERE id=$2

db.namepayments

db.userpayments_rw

server.addresspg-primary.internal:5432

~/payments-api/endpoints/capture

prodlive

POST/v1/payments/:id/capture

Servicepayments-api

Requests8.1k

Rate22 /min

Latency86 ms p50 5.01 s p95

Error rate · 5xx4.2%

500340

Warn rate · 4xx0.4%

42229

Endpoint health

requestsp50p95error rate

09:0010:3012:0013:3015:00

Recent errors

StatusErrorTraceAge

500UpstreamTimeoutError connect ETIMEDOUT acquirer-gw.internal:4437a3f9c1e2 m

500UpstreamTimeoutError connect ETIMEDOUT acquirer-gw.internal:443b81d02aa9 m

422ValidationError amount must be positive4c77e0f131 m

Get alerted when errors arise, monitor their occurrences over time, and resolve them as they’re fixed.

Docs Node.js Python

Automatic insights

Detect anomalies in your running services.

Vigilon compares error rates, latency, and traffic with recent history and alerts you if an anomaly is detected.

payments-api · prod

Error rate spike detection

See when a service’s error rate rises above its usual level.

POST /v1/payments/:id/capture8.1k req · clears gate

Latency regressions

See when an endpoint becomes slower than usual.

POST /v1/webhooks/acquirer

Traffic drops

Find unexpected drops in request traffic.

# 36 m ago
service   payments-api · prod
change    error rate 0.3% → 8.8%
window    last 30 m vs previous 24 h
deploy    v2.3.2 · near spike start

Evidence behind insights

Each finding shows what changed, where, and when.

Background jobs

Monitor scheduled jobs and queue workers.

Track jobs alongside your services.

Missed runs: Flag jobs that do not run on schedule.
Traces for every run: Open a failed run and inspect its trace.
History and duration: Compare failure rate and p50/p95 runtime by job.

Docs Node.js Python

~/payments-api/jobs

prodlive

Runs

3.8ktotal

Failure rate

1.31%

50failed

Failing

1job

Overdue

1job

Monitored

5jobs

Jobs

JobServiceScheduleHealthLast runRunsFailp50p95

sync-exchange-ratesexchange-rates-job*/5 * * * *Healthy2 m ago2.0k0.05%184 ms412 ms

webhook-retrybilling-worker—Failing4 m ago1.0k4.8%231 ms604 ms

nightly-reportreporting-worker0 3 * * *Overdue2 d ago50%18.7 s24.1 s

invoice-rollupinvoice-rollup-etl0 * * * *Healthy14 m ago1680%640 ms1.19 s

send-emailemail-service—Healthy9 m ago6720%41 ms88 ms

MCP for coding agents

Bring production context into your coding session.

Connect Vigilon to MCP clients. Your agent can inspect errors, traces, and endpoint health while it works in your repo.

Why did POST /webhooks/stripe start returning 500s after today’s deploy?

vigilon · list_error_groups→ SignatureVerificationError · 87 events

vigilon · get_trace 8dc4a12f→ exception in webhook.ts:84

The failures all come from signature verification. Today’s deploy parses the request body before Stripe sees it, so the raw payload is gone. The change is in webhook.ts. Want me to patch it?

01
Add the server
Add the MCP server to your coding agent.claude mcp add --transport http vigilon https://mcp.vigilon.io/mcp
02
Ask about production
Ask your agent to investigate a bug that showed up in production or dig into an insight about a regressing API endpoint.
03
Ship the fix
Watch your agent do the work you normally would.

OpenTelemetry

Send standard OpenTelemetry data.

Point any OTLP exporter at Vigilon. Run it beside your current backend. Keep your existing instrumentation.

Docs Use an existing OpenTelemetry setup

Time to first signal

It’s so easy to get started.

We recorded the flow from signup to live telemetry in one take, as proof.

Signup to first signal

01
Signup
Create your workspace and start your plan.
02
Create a project and API key
Name the project after your product, then generate an API key under its settings page.
03
Run the installer in your AI agent
Run `curl -fsSL https://vigilon.io/agents/install.md` and follow the instructions in that file to set up Vigilon in this project.
04
Add your API key
VIGILON_API_KEY=[YOUR_API_KEY]

First signalorders-api · GET /api/orders · 200 · 142 ms

Early access

Join the early-access waitlist.

We’re looking for users who can use the product and give us feedback. Everyone on the waitlist gets 2 months of access to Vigilon for free.

Everything you need to watch production.

Detect anomalies in your running services.

Error rate spike detection

Latency regressions

Traffic drops

Evidence behind insights

Monitor scheduled jobs and queue workers.

Bring production context into your coding session.

Add the server

Ask about production

Ship the fix

Send standard OpenTelemetry data.

It’s so easy to get started.

Signup

Create a project and API key

Run the installer in your AI agent

Add your API key

Join the early-access waitlist.