Model
Define dimensions, measures, joins, calendars, hierarchies, glossary terms, and model versions once.
Governed semantic layer for BI, Excel, and AI
Define metrics once. Use them safely everywhere.
Tessallite sits between your BI tools and source data. It maps raw tables to approved business definitions, enforces access rules, routes each query through the fastest safe path, and gives dashboards, spreadsheets, APIs, and AI agents the same trusted answer.
Semantic modelling
JDBC and XMLA endpoints
Row and column security
Automatic query acceleration
Product category
A semantic layer and query acceleration platform.
Tessallite gives platform teams one controlled layer for metric definitions, data access, query routing, and business-facing metadata. To the tools, Tessallite looks like a database. To the business, it becomes the place where important numbers have one approved meaning.
Govern
Apply workspace roles, project access, personas, row security, column security, and data tags consistently.
Route
Serve each query through an aggregate, pocket table, or source path depending on what is safe and accurate.
Explain
Expose owners, freshness, lineage, route details, query logs, and model health so teams can trust the answer.
Product proof
Use real product surfaces.
Tessallite leads with the working surfaces evaluators need: the modeller canvas, conversational agent, query routing, governed endpoints, and Excel reporting. The full Product Tour describes what each surface does and when to use it.
Conversational agent
Ask in plain business language and get a written answer, a chart, the underlying rows, and the semantic query — all from the governed model.
Query routing
Shows parse, bind, route, rewrite, and result steps so teams can see when a query used an aggregate instead of the source path.
Excel workflow
Lets analysts work from familiar workbook and PivotTable patterns while still using governed model definitions.
Enterprise control
Security rules travel with every answer.
Tessallite does not treat governance as a separate brochure page. The same access checks apply when a user asks through Excel, BI, SQL, an API, or the conversational agent.
Audience-aware models
Personas and roles shape which measures, dimensions, hierarchies, and filters each audience can use.
Row and column controls
Row security, data tags, and column restrictions prevent sensitive fields from leaking through another client.
Audit and trace
Query logs, route badges, owners, freshness, and lineage give reviewers evidence for where numbers came from.
Performance
Fast dashboards without hand-maintained summary tables.
Tessallite observes repeated query patterns, builds useful aggregates, keeps them refreshed, and falls back to the original source when a prepared answer is not safe for the query shape.
Aggregates
Reusable rollups for common grains and measures reduce repeated scans over large fact tables.
Pocket tables
Controlled slices support repeated filtered analysis without opening access to the entire source shape.
Source fallback
If no safe match exists, Tessallite executes through the source path and records the miss for later optimisation.
Connected tools
Meet users in the tools they already use.
Tessallite exposes governed models through standard endpoints: JDBC on the PostgreSQL wire protocol, XMLA for Excel and Power BI, REST for application access, and agent context for natural-language questions.
Excel and Power BI
Connect through XMLA, browse cubes, use friendly field names, and refresh workbooks against the deployed model.
SQL and BI tools
Use PostgreSQL-compatible clients such as DBeaver, Power BI PostgreSQL, Tableau, pgAdmin, or custom drivers.
Apps and agents
Use headless APIs, embedded chat, and the conversational client when the answer needs to live inside another workflow.
PostgreSQL
BigQuery
Spark
Redshift
Operational readiness
Built for model owners after launch.
The product includes model health, query logs, usage analytics, impact analysis, schema-change review, refresh history, alerting, and export/import workflows so the semantic layer can be operated, not just designed.
Model health
Find invalid dimensions, broken measures, failed aggregates, refresh issues, and warnings before users hit them.
Usage and impact
See which questions, reports, and columns are used so model changes can be reviewed with evidence.
Export and import
Move model and project definitions without copying source credentials, then rebind them in the target environment.
Evaluation to production
Start locally. Move to supported rollout when the case is proven.
Community is for a self-hosted evaluation with the full product surface and practical limits. Enterprise removes operating limits and adds deployment options, support, SSO, and security evidence for production teams.
Community
Free local evaluation with the same modelling, BI, agent, governance, and acceleration surface.
Install guide
Follow the Docker Compose path, verify the signed bundle, add the licence, and open the local UI.
Enterprise
Use Enterprise for unlimited tenants, models, users, SSO, support SLA, and production deployment guidance.
Give every team one governed way to ask for numbers.
Book a product walkthrough or start a Community evaluation with the local bundle and public help path.