Connect and model
Diverse sources, reproducible transformations, and portable semantics that define what each metric means.
Product vision · in development
The alternative to Power BI and Tableau. Built for agents.
DataVIZ envisions a complete business intelligence platform: connected data, a shared semantic model, visual analysis, and agents working from the same governed knowledge base.
The goal is to replace the full BI workflow, not just translate dashboards. The platform is still in development.
DataVIZ architecture One source of truth
From data to action,
without losing context.
Inputs
Data sourcesPower BITableauProcesses
Shared coreEnterprise knowledge
Experiences
Metrics · relationships · policies · lineage · decisions
PeopleExplore, create, and decide.
AgentsQuery, reason, and propose.
Permissions, evidence, and human oversight at every layer
Dashboards show answers. The enterprise needs to retain the knowledge behind them.
A complete platform
Your own BI, from model to decision.
The goal is to replace the Power BI and Tableau experience end to end: connect sources, define metrics, build analyses, share results, and operate with traceability. Migration is an entry point, not the destination.
Explore and create
Dashboards, filters, visuals, and interactions in a dedicated web runtime, with authoring for people, SDKs, and agents.
Trust and operate
Provenance, permissions, lineage, validation, and audit trails so every result has context and accountability.
Enterprise knowledge base
Shared memory for people and agents.
DataVIZ aims to bring definitions, relationships, rules, and evidence into a queryable core. A metric becomes more than a number in a report: it carries meaning, origin, permissions, and a decision history.
The same semantic layer should power dashboards, search, and agent actions without rebuilding context in every tool.
DefinitionsMetricsWhat is measured and how it is calculated.
ContextRelationshipsHow data and processes connect.
DataVIZGoverned knowledgeA shared, versioned, verifiable foundation.
ControlPoliciesWho may view, propose, or approve.
TrustEvidenceWhere every answer comes from.
Agentic from day one
Agents are part of the architecture, not a chatbot added later.
They should work with the same concepts, permissions, and evidence as people. They can assist with analysis and authoring while sensitive changes remain reviewable and subject to approval.
01
Query with meaning
Find data, metrics, and policies with semantic context and provenance.
02
Reason within boundaries
Use permitted tools and data; show evidence and acknowledge uncertainty.
03
Propose changes
Prepare analyses, models, or visuals as typed changes a person can review.
04
Act with oversight
Apply only authorized actions, with traceability, permissions, and human approval where needed.
Transparent progress
An ambitious vision, built in stages.
DataVIZ is not yet a production-ready replacement. Current work demonstrates technical components; the complete platform experience remains the roadmap goal.
In development
BI translation and validation
Technical Tableau → Power BI MVP with fidelity checks. It preserves intent and surfaces migration gaps.
Next
DataVIZ web runtime and authoring
Viewing, exploring, and creating dashboards on portable semantic contracts.
Vision
Enterprise-scale knowledge and agents
Catalog, lineage, policies, and agents integrated into a governed, deployable platform.