✦ VectorData is live · Open workspace →

Product

The calm platform for BigQuery & dbt reliability

Monitoring. Data CI triage. Blast radius. AI insights. Calm alerts. Warehouse FinOps. One command center — built like a mature platform, sized for lean teams.

vectornosis.com/app

Northwind Outfitters — Dashboards

sample
UploadLive · checked 3s ago
$project:northwind$env:prod$region:us-central1$team:data-platform$service:dbtPast 1h

health

88

models

13/16

tests

7/10

failed

3

incidents

4

fresh.sli

63%

bq.mtd

$4,203

HEALTH 88MODELS 13/16FAILED 3BQ MTD $4,203

WAREHOUSE

prod · us · data · bq · 1h

SLO · Freshness uptime

99.12%

target 99% · 7d / 90d

SLO · dbt build success

97.4%

target 98% · 3 failed this window

Vector · Watchdog

Schema drift on fct_orders — 3 noisy failures → 1 calm alert

Root cause · blast 2 marts + exec KPI · est. risk ~$48k/week if leadership trusts the board

Open incident

Calm alert routing

Naive tooling

12

Calm routing

1

One root-cause page to #data-alerts — child pages muted 2h

Monitors status

Alert 2Warn 2OK 52No Data 1

Grouped child pages into one root-cause alert

Service map · DAG

Fail-first · 12 nodes

fct_orders → fct_revenue_daily → mart_exec_kpis

Models built / total

13/16

Error / fail rate

0.7%

Avg build latency

96.8s

BQ MTD spend

$4.2k

Live warehouse telemetry · Past 1h

Freshness lag (hours)

Query errors

BQ $/hour

Pipeline & freshness

Table health map

orders · critical

Event stream

  • CRITstg_orders · 3 tests failed
  • WARNorders lag 26h / 24h SLA
  • WARNschema drift · fct_orders
  • OKcost back under budget

Fix first

orderscritical

Last update 3 days ago · SLA 24h

Open workspace →

Same chrome as /app — try the sidebar, tiles, or Replay

Open workspace →

The platform

See it all in one place — for BigQuery & dbt

Datadog-style product breadth. Metaplane-class reliability jobs. A lean wedge you can actually operate: monitoring, lineage impact, data CI, insights, alerts, cost, and trust — with Live / Partial / Soon labeled honestly.

Monitoring

Live

Freshness SLAs on BigQuery and Snowflake, dbt run/test health, and BigQuery cost baselines in one command center.

  • ·BQ + Snowflake freshness thresholds
  • ·dbt model + test failures
  • ·BigQuery job cost baselines & spikes
Explore →

Lineage & blast radius

Live

See which KPIs and models sit downstream of a break — impact score, Fix-first, OpenLineage export.

  • ·Downstream impact on incidents
  • ·Fix-first ranking by blast radius
  • ·OpenLineage export — metadata isn’t hostage
Explore →

PR Data CI / circuit breaker

Live

GitHub Action posts Check Runs + PR comments with Fix-first — fail the PR / block merge when critical models fail.

  • ·action/dbt-ci + example workflow
  • ·Circuit-breaker: fail Check on bad dbt
  • ·Optional block-merge policy for critical models
Explore →

AI Watchdog insights

Live

Plain-language triage: what broke, who feels it, what to fix first — not anomaly theater.

  • ·Unified Fix-first queue
  • ·AI root-cause summaries
  • ·Leadership-ready incident context
Explore →

Automated calm alerts

Live

One notification per root cause. In-app + email Live; Slack / PagerDuty stay Coming Soon until the integrations wave.

  • ·Root-cause grouping (Live)
  • ·Email + in-app (Live)
  • ·Slack · PagerDuty (Coming Soon — 3P)
Explore →

Incident lifecycle

Live

Assign, status, mute, impact score, audit trail, one-click blameless postmortem — not Slack-only paging.

  • ·Declare → assign → resolve
  • ·Impact score × exec KPI $
  • ·Auto-draft postmortem notebooks
Explore →

Flat-Price Guardian

Live

Your Vectornosis invoice is flat. Guardian watches BigQuery pace vs baseline and warns before finance sees the bill.

  • ·Month-end spend forecast
  • ·Hard cap thresholds
  • ·Kill-switch recommendations on spike jobs
Explore →

Business / semantic checks

Live

Catch wrong WHERE filters and broken revenue logic — green anomaly monitors are not enough.

  • ·SQL assertion monitors
  • ·Row-count & KPI contracts
  • ·Trust score when business checks pass
Explore →

OpenLineage & portability

Live

Export runs and lineage as OpenLineage events. Leave without rewriting the org.

  • ·/api/openlineage emit + export
  • ·No proprietary metadata prison
  • ·Works with dbt artifacts you already have
Explore →

Warehouse FinOps

Live

Catch 2× scans and missing partition filters before finance sees the invoice.

  • ·Spike detection vs baseline
  • ·Top expensive jobs
  • ·Playbooks for common fixes
Explore →

Command center views

Live

Mission control for lean teams — freshness, dbt, cost, and Watchdog in one calm surface.

  • ·Empty-first workspace + sample Import
  • ·Metric strip · DAG · calm routing · Watchdog
  • ·Upload run_results.json or load sample
Explore →

Security & independence

Live

Metadata only. Independent company — not Datadog-owned. Least-privilege SA; rows never leave your warehouse.

  • ·BQ INFORMATION_SCHEMA + JOBS
  • ·Independent roadmap (post–Metaplane acquisition)
  • ·Viewer / Analyst / Admin RBAC
Explore →

01 · Monitoring

Live

See warehouse reliability in one place

Like a mature observability platform — multi-warehouse freshness (BigQuery + Snowflake) plus the dbt and cost signals lean teams operate every morning.

  • Freshness SLAs on BigQuery INFORMATION_SCHEMA and Snowflake LAST_ALTERED
  • dbt run and test status beside warehouse truth
  • BigQuery cost baselines so reliability and spend share one ritual
  • Explore the sample board before you share credentials
Open command center →

See warehouse reliability in one place

Live
Command centerLIVE
Freshness99.12%
dbt build13/16
BQ MTD$4.2k

02 · Data CI / CD

Live

Block bad dbt in the pull request

Metaplane’s PR Data CI/CD story is now a VectorData Live path: GitHub Check Run, PR comment with Fix-first, job fails on model/test failures.

  • Composite Action under action/dbt-ci
  • Uploads run_results.json to /api/webhooks/dbt
  • Posts Check Run + PR markdown with Fix-first
  • Works with dbt-bigquery or dbt-snowflake artifacts
PR Data CI →

Block bad dbt in the pull request

Live

run_results.json

stg_orders3 tests failed
fct_revenuemodel errored
customer_ltvupstream blocked

Fix first: repair stg_orders

03 · Lineage & impact

Live

Know the blast radius before the war room

Impact analysis is how mature tools prioritize. VectorData shows downstream KPIs, impact scores, and OpenLineage export so metadata isn’t hostage.

  • Incident cards include downstream context
  • Fix-first ranking by blast radius + exec $
  • OpenLineage export for portable lineage
  • KPI-linked blast radius on the service map
Blast radius →

Know the blast radius before the war room

Live

Blast radius

  1. 1.orders · stale
  2. 2.fct_revenue · downstream
  3. 3.exec KPI · impact

Partial today · deeper lineage Soon

04 · Data insights

Live

AI that explains the break — then gets out of the way

Data insights should answer 7am questions: what broke, who feels it, what to fix first. That is Watchdog — not a new category slogan.

  • Unified incident queue across freshness, dbt, and cost
  • Plain-language AI summaries grounded in metadata
  • Suggested next action on the critical path
  • Shareable context for standup and exec threads
AI Watchdog →

AI that explains the break — then gets out of the way

Live

AI · Watchdog

Schema drift on fct_orders

AI root cause · blast radius → exec KPI · est. risk if stale

Open incident → Fix-first queue

05 · Automated alerts

Live

Calm routing beats another Slack storm

Mature alerting is rarity + clarity. VectorData groups by root cause; email is Live; Slack/PagerDuty stay Coming Soon as third-party apps.

  • One notification per root cause (Live)
  • In-app queue + email delivery (Live)
  • Mute windows + incident assign/status (Live)
  • Slack · PagerDuty Coming Soon (3P)
Calm alert routing →

Calm routing beats another Slack storm

Live

Calm alert

orders stale 3 days · 1 notification · 3 downstream KPIs

Not 12 Slack pings for child models

06 · Cost & FinOps

Live

Stop BigQuery bill shock with playbooks

Warehouse FinOps belongs next to reliability. Spike detection, job identity, and partition-filter playbooks ship in the same command center.

  • Flag jobs that run 2×+ vs baseline
  • Surface top expensive queries with owners
  • Pair with the free cost calculator
  • Keep monitoring on a flat plan — no second surprise invoice
Cost spikes →

Stop BigQuery bill shock with playbooks

Live

Flagged query

$1,940

2.4×

missing partition filter

07 · Dashboards & collaboration

Live

A shared reliability view — not twenty private tabs

Metaplane’s dashboards story is about shared truth. VectorData’s command center is the lean-team version: one calm surface stakeholders can open.

  • Health tiles for freshness, dbt, and spend
  • Event stream + Fix-first queue
  • Load sample board for demos and onboarding
  • Same URL for engineering and operators
Open workspace →

A shared reliability view — not twenty private tabs

Live
Freshness
dbt
Cost
Watchdog

08 · Trust layer

Live

Metadata only — security review that can finish

Mature platforms earn trust with a clear access model. VectorData never needs your rows to tell you a table is stale.

  • INFORMATION_SCHEMA, JOBS, dbt artifacts
  • Least-privilege service account
  • Public security page for reviewers
  • Revoke access anytime
Security model →

Metadata only — security review that can finish

Live

Access model

Table timestamps
Job statistics
dbt artifacts
Never row samples

Breadth with honesty

Built to win the bake-off — not apologize in the footnotes

PR Data CI/CD and multi-warehouse freshness (BigQuery + Snowflake) are Live. Metaplane may still go wider on deep lineage catalogs and anomaly sprawl — VectorData wins the lean-team morning check and the PR gate.

AreaMetaplane-class jobVectorData todayStatus
MonitoringBroad warehouse monitors + anomaliesFreshness + dbt + cost — BigQuery & SnowflakeLive
Multi-warehouseMany warehouses out of the boxBigQuery + Snowflake freshness (Live); more adapters nextLive
Lineage & impactDeep lineage / impact analysisBlast radius on incidents; deeper maps SoonPartial
Data CI/CDPR-time data quality workflowsGitHub Check Run + PR comment + fail-on-fail (Live)Live
Data insightsUsage & stack insightsAI Watchdog triage insights (Live)Live
Automated alertsMulti-channel alertingCalm root-cause alerts; more channels SoonPartial
DashboardsCustom DQ dashboardsCommand center + shared demo views (Live)Live
Schema driftSchema change alertsdbt tests Live; dedicated drift SoonPartial
Job monitoringPipeline latency / jobsBigQuery JOBS cost + duration pathLive

Prefer a bake-off? VectorData vs Metaplane → or vs Datadog →

Security

Metadata only. Rows never leave BigQuery.

Table timestamps, dbt artifacts, job statistics — never row samples. Least-privilege service account. Revoke anytime.

Read the security model →

Start with the live product

Open the command center, or go deeper on VectorData.