About InsightViz

InsightViz — built by VerticalServe

AI-native business intelligence for teams leaving Tableau and Power BI. Part of the InsightLake enterprise AI suite by VerticalServe.

Our story

From migration engagements to a product

For years VerticalServe has helped enterprises move data platforms to the cloud. The BI layer was always the part that stayed behind: hundreds of Tableau workbooks nobody could review, metric definitions that disagreed, viewers paying for seats to look at a chart. InsightViz is what we built once the pattern was clear enough to productise.

2024

The pattern

Every migration ended with the same three asks: keep the dashboards, make the metrics agree, and let people ask their own questions without a ticket.

  • Dashboards as reviewable documents
  • One semantic model per source
  • AI grounded in that model, not in tables
2025

The engine

DuckDB over Parquet extracts in the customer's own bucket turned out to answer both the performance question and the licensing one.

  • Versioned extracts, atomic flips
  • Millisecond queries on the server
  • Browser and desktop runtimes on the same contract
2026

The product

InsightViz ships on the shared VerticalServe platform: one sign-in, one connection registry, one design system across the InsightLake suite.

  • Tableau import with a fidelity report
  • Row/column policies compiled into every query
  • A query log that answers the auditor
What we believe

Product principles

The decisions that shape every feature.

Everything is a text file

Dashboards, models and charts are JSON and YAML. If it cannot be diffed, reviewed and promoted like code, it does not belong in the product.

The model is the contract

Dashboards, Explore and the AI all read the same semantic model. A metric means one thing everywhere, and changing it is one change.

Trust comes from transparency

Every answer shows the query it ran, the SQL it compiled and the policies it applied. We do not ask people to trust accuracy claims.

Client-side is UX, not security

Data that must not leave the server never does. Anything served to a browser or desktop is data that person may see, full stop.

Your cloud, your keys

Extracts live in your S3, GCS or Blob. Credentials are secret refs, never values. The model provider is yours to choose.

AI for the boring parts too

Deriving a model from a schema, naming 300 columns, translating calculated fields, suggesting the questions worth asking.

By the numbers

Built on real enterprise workloads

30+

Products in the InsightLake suite

50+

Enterprise deployments via VerticalServe

10+ yr

Data & AI engineering pedigree

3 clouds

AWS, GCP, Azure certified

Talk to the team that built it

Migration scoping, a demo on your workbook, or a conversation about where BI is going.

Request a demo