Bronto
Integration
Bronto
Vercel

Bronto for Vercel

Stream every log, trace, and metric from your Vercel deployments into Bronto. Search across all of it in sub-seconds, keep it hot forever, and investigate incidents with AI instead of grepping.

Book a DemoSign Up Free
Why Bronto + Vercel

Full-stack observability built for the AI era, wired straight into Vercel

Bronto is a full-stack intelligent observability platform built on BrontoDB, orders of magnitude more efficient than legacy tools. No more choosing between blindspots and overstretched budgets, especially as AI workloads push data volumes 10 to 100x higher.

Bronto dashboards showing Vercel traces, Web Analytics, and Speed Insights
Better outcomes
4-in-1

All Vercel drains, one home

Logs, Traces, Speed Insights, and Web Analytics in one place. Correlate Core Web Vitals with backend traces.

12mo

Always-hot retention

A full year of Vercel telemetry queryable in milliseconds. No cold tiers, no rehydration.

100x

More cost efficient

Built for AI-era data volumes, without the legacy hot-tier bill.

<1s

Sub-second search

Query TBs and PBs of Vercel logs and traces in milliseconds, consistently.

1-Click

End-to-end correlation

Jump from trace to logs to code in seconds.

Faster

MTTR with AI

Every alert ships with an investigation report. Bronto Scope walks you straight to the root cause.

Drain coverage
  • LogsRuntime logs from your Vercel deployments
  • TracesDistributed trace spans across your serverless functions
  • Speed InsightsCore Web Vitals and real user performance metrics
  • Web AnalyticsVisitor events and page view data
AI-native parsing

Parsed for AI, by default

Bronto auto-parses Vercel AI SDK and LLM traffic to the OpenTelemetry GenAI semantic conventions and OpenLLMetry. Fields like gen_ai.system, gen_ai.request.model, gen_ai.usage.*, prompts, completions, latency, and cost arrive structured and queryable without manual instrumentation.
OpenTelemetry GenAI conventions
Vercel logs → OTEL GenAI fields
  • Raw log
    [POST] /api/chat 200  fn=us-east-1::chat  dur=1284ms  "model=gpt-4o tokens_in=842 tokens_out=311 cost=$0.0089"
    Parsed
    gen_ai.system=openaigen_ai.request.model=gpt-4ogen_ai.usage.input_tokens=842gen_ai.usage.output_tokens=311gen_ai.usage.cost_usd=0.0089duration_ms=1284
  • Raw log
    [POST] /api/agent 200  fn=iad1::agent  dur=3421ms  "anthropic claude-3-5-sonnet in=1204 out=892 $0.0276"
    Parsed
    gen_ai.system=anthropicgen_ai.request.model=claude-3-5-sonnetgen_ai.usage.input_tokens=1204gen_ai.usage.output_tokens=892gen_ai.usage.cost_usd=0.0276duration_ms=3421
  • Raw log
    [POST] /api/embed 200  fn=fra1::embed  dur=187ms  "model=text-embedding-3-small tokens=2048"
    Parsed
    gen_ai.system=openaigen_ai.request.model=text-embedding-3-smallgen_ai.usage.input_tokens=2048gen_ai.operation.name=embeddingsduration_ms=187
Auto-parsed at ingest · no manual instrumentation
Agentic observability

An observability platform agents can actually use

Bronto is designed for both humans and agents. Plug Vercel telemetry into your AI SRE, copilot, or vibe coding workflow without rebuilding your stack.

  • MCP server out of the box

    Claude, Cursor, the AWS DevOps agent, and any MCP-compatible agent can query Vercel logs, traces, and metrics directly.

  • Cost-aware by design

    Up to 100x cheaper than legacy hot tiers.

  • Context enrichment

    Statement Patterns and Statement IDs group similar logs and link back to the exact code path, so agents reason over structure instead of raw text.

  • Structured, not stringy

    Every field parsed at ingest, so agent tool calls return typed JSON instead of regex guesses.

  • OTEL GenAI + OpenLLMetry support

    Agentic and non-agentic workloads in one place, at production scale.

Logs
Metrics
Traces
Intelligence layer
What you get

Everything you need to operate apps on Vercel

Bronto log search in dark mode

Streaming runtime logs

Every function invocation and edge log, no sampling required.

Bronto distributed traces in dark mode

Distributed traces

Span-level visibility across serverless and edge functions, linked to the originating log.

Bronto infrastructure metrics dashboard

Speed Insights and Web Vitals

LCP, INP, and CLS for real users, alongside your backend data.

Bronto sub-second log search

Sub-second search at any scale

BrontoDB returns queries in milliseconds, across petabytes of telemetry.

Bronto monitors and alerts in dark mode

Monitors and alerts

Anomaly detection and threshold monitors. Every alert ships with an investigation report.

Bronto dashboards in dark mode

Dashboards on your data

Build views by query, or vibe your own UI with BrontoVibe.

One-click correlation

Built-in AI troubleshooting that slashes your MTTR

Stop hopping between tabs and grepping through logs. Jump from a Web Vital regression to the trace, to the exact log line, to the offending function in a single click. Every alert ships with an AI-generated investigation report so you know what went wrong, where, and why, before you finish your first sip of coffee.

  • Trace, logs, and code linked end-to-end across your Vercel stack.
  • AI investigation reports attached to every alert, no prompting required.
New ways of working

Optimised for your new workflow

Whether you live in your IDE, your terminal, or your own UI — Bronto plugs into the way you already work.

Claude Code asking Bronto MCP about latency spike5xx summary report from Bronto MCPCDN request volume spike analysis

Bronto MCP

Plug Bronto straight into Claude, Cursor, and any MCP-aware agent so your logs, metrics, and traces become a first-class tool call.

Learn moreRead blog

REST & Query API

Programmatic access to every query, dashboard, and alert.

CLI & Terminal

Search, tail, and pipe Bronto data right from your shell.

Build a live error rate table grouped by service, with a sparkline for the last 24 hours beside each row.Create a deploy tracker that highlights any spike in error logs within 10 minutes of a new deployment.Build a cost dashboard that shows my ingestion volume by team or project so I can see who's logging the most.Create an alert inbox where I can snooze, acknowledge, and add notes to firing alerts without leaving the UI.Build a trace waterfall viewer that lets me search by trace ID and see every span with its duration and status.Create a side-by-side log diff tool so I can compare log patterns from this week versus last week.Build a customer health scorecard that shows error counts and latency per customer ID in a sortable table.Create a terminal-style log tail UI that lets me filter by severity and keyword in real time.Build a live error rate table grouped by service, with a sparkline for the last 24 hours beside each row.Create a deploy tracker that highlights any spike in error logs within 10 minutes of a new deployment.Build a cost dashboard that shows my ingestion volume by team or project so I can see who's logging the most.Create an alert inbox where I can snooze, acknowledge, and add notes to firing alerts without leaving the UI.Build a trace waterfall viewer that lets me search by trace ID and see every span with its duration and status.Create a side-by-side log diff tool so I can compare log patterns from this week versus last week.Build a customer health scorecard that shows error counts and latency per customer ID in a sortable table.Create a terminal-style log tail UI that lets me filter by severity and keyword in real time.
Build an on-call dashboard that surfaces the top 5 most-fired alerts from the last 7 days with a count and owner.Create a latency heatmap across my services so I can see which endpoints are slowest at which times of day.Build a log volume chart that sends a Slack message when any service crosses a threshold I set.Create a feature flag impact viewer that shows log error rates before and after a flag was toggled.Build a multi-tenant log explorer where I can switch between customer environments with a single dropdown.Create a status page builder that auto-populates uptime percentages from my trace and log data.Build a root cause timeline that stitches together related log events into a single chronological incident view.Create a daily digest email that summarises my top errors, slowest endpoints, and ingestion volume from the past 24 hours.Build an on-call dashboard that surfaces the top 5 most-fired alerts from the last 7 days with a count and owner.Create a latency heatmap across my services so I can see which endpoints are slowest at which times of day.Build a log volume chart that sends a Slack message when any service crosses a threshold I set.Create a feature flag impact viewer that shows log error rates before and after a flag was toggled.Build a multi-tenant log explorer where I can switch between customer environments with a single dropdown.Create a status page builder that auto-populates uptime percentages from my trace and log data.Build a root cause timeline that stitches together related log events into a single chronological incident view.Create a daily digest email that summarises my top errors, slowest endpoints, and ingestion volume from the past 24 hours.

Vibe it with v0

Bronto's UI is open source. Fork it as a v0 template, prompt your way to custom dashboards and internal tools on top of the Bronto API, and ship straight to Vercel.

Learn moreRead blog

AI SREs

Let autonomous agents triage incidents, investigate anomalies, and surface root cause in seconds.

Your IDE & Coding Agents

Query your telemetry without leaving your IDE or coding agent of choice.

Get started

Up and running in minutes

  1. 1

    Create a Bronto account

    Sign up free and generate an API key with the Ingestion role.
  2. 2

    Open Vercel Drains

    Go to Team Settings → Drains → Add Drain. Requires Vercel Pro or Enterprise.
  3. 3

    Point the drain at Bronto

    Select Custom Endpoint and use https://ingestion.<region>.bronto.io/.... Add the three required headers below. Repeat for each drain type you want to enable.
  4. 4

    Verify data in Bronto

    Open your dataset and watch telemetry flow in. Done.
  5. 5

    Use pre-built templates

    Run Bronto Express to instantly provision dashboards, monitors, and saved searches that match the Vercel drains you've connected. Re-run it anytime you add a new drain.
Required headers
x-bronto-api-key:    <your-api-key>
x-bronto-dataset:    <dataset-name>
x-bronto-collection: <collection-name>
Pre-built templates (coming Late May)

Pre-built dashboards in one click

Bronto Express detects which Vercel drains you've connected and provisions the matching dashboards instantly. Each dashboard maps to a different drain type, so you can ship with any subset of Logs, Speed Insights, and Web Analytics.

View in docs
Vercel Overview dashboard previewVercel Web Analytics dashboard previewVercel Web Vitals dashboard preview
1 / 3
Vercel OverviewLogs drain

High-level view of request traffic, error rates, cache behaviour and firewall activity across your Vercel deployments.

  • Lambda Error Spike
    Logs drain
    Critical
  • Lambda Crash
    Logs drain
    Critical
  • 5xx Error Rate Spike
    Logs drain
    Critical
  • High 4xx Rate
    Logs drain
    Warning
  • Firewall Deny Spike
    Logs drain
    Warning
  • LCP P75 Regression
    Speed Insights drain
    Warning
  • INP P75 Regression
    Speed Insights drain
    Warning
Lambda Errors
Logs drain

Error and fatal log events from lambda sources

Lambda Crashes
Logs drain

Events with status code -1 (process crash)

Firewall Denies
Logs drain

Requests denied by the Vercel WAF

Build Failures
Logs drain

Error-level events from the build source

Poor LCP Pages
Speed Insights drain

LCP measurements above 2,500 ms

Poor INP Pages
Speed Insights drain

INP measurements above 200 ms

FAQ

Common questions

Give your Vercel logs a better home

Sub-second search, 12 months hot retention, and AI-native parsing for every drain.

Start free on VercelBook a Demo