Skip to content

All ears for your numbers.

Live Analysis for Power BI

Fennibi Ask your Power BI models. Answered live.

Fennibi, formerly SemaDoc · About the new name →

Fennibi answers your question with real figures: the assistant picks the right Semantic Model, plans the analysis in several steps, runs read-only DAX queries inside your own Fabric capacity, and brings the result together as charts, tables, and a conclusion backed by the numbers.

For that to work, the assistant needs a model it can understand. That is exactly what Fennibi delivers: complete, fully automated documentation of every Semantic Model — the foundation of every good answer, and your searchable data catalog at the same time.

What Makes a Good Answer From a Power BI Model

AI assistants rarely fail because of the question — they fail because of what they find inside the model. Three things decide whether a business question turns into an answer you can rely on.

Answers that are factually right

An assistant is only as precise as the model is described. When tables, columns and KPIs carry business names and explanations, it picks the right measure instead of one that merely sounds plausible. Fennibi generates exactly that description automatically for every Semantic Model.

Answers while the question still matters

Otherwise every new question needs a new visual or a ticket to the BI team. On a documented model the assistant finds its own way to the figure and answers in the conversation — in minutes instead of days.

Figures everyone reads the same way

Every report author presents differently, and every analysis costs reading time all over again. Fennibi answers in one consistent notation based on IBCS®: the same rules for actual, plan, prior year, and variance — in every single answer.

Fennibi: Live Analysis on Documented Power BI Models

You ask your question in plain language. Fennibi picks the right Semantic Model, takes the figures straight from your Power BI model and shows how they came about. They stay your figures — no second version of the truth, no Excel export on the side.

  • ✓ Answers with real figures from your model — traceable down to the individual query
  • ✓ Your data stays with you: everything is calculated read-only inside your own Power BI or Fabric environment
  • ✓ Everyone sees exactly what they are allowed to see in the report — existing permissions apply unchanged
  • ✓ Dependable calculations: comparisons, trends and forecasts follow fixed rules and are repeatable at any time
  • ✓ Results as chart and table in one consistent notation — readable at a glance
  • ✓ The foundation: a searchable catalog of all your Semantic Models — documented fully automatically, versioned, in 14 languages

Live Analysis: Answers Straight From Your Power BI Model Preview

Ask your question in plain language. Fennibi plans the analysis, runs it against your own Semantic Model, and backs every figure with the query it came from.

Calculate instead of guess

“How did revenue per region develop over the last six months?” — the assistant plans the steps, runs read-only DAX queries against your model, and shows the underlying query next to every result.

Follow up without starting over

“And without the three largest customers?” — the conversation keeps its context, and figures already retrieved are reused instead of queried again.

Results you can read

Charts and tables follow the same notation throughout, based on the IBCS® rules: actual, plan, and prior year clearly distinguishable, deviations shown as variance, one scale per group.

Movement and forecast

Who joined, who dropped out, who grew, who shrank? Fennibi computes such roll-ups and simple projections deterministically in Python — over the complete result set, not over a sample.

Find the right model

“Which model contains revenue per region?” — the assistant finds the matching Semantic Model in your catalog before it starts calculating.

The analysis as a PDF report

Question, analysis steps, queries, charts, and conclusion in a single document — ready to share and to archive.

To try it out, Fennibi Docs is enough: sign in on fennibi.com with your Microsoft account and query your own model right in the browser. In the locally installed Fennibi, live analysis is part of the Fennibi subscription.

Preview: we keep working on live analysis. What is there today already delivers dependable results — and every number stays traceable through the DAX query it was computed with.

IBCS® is a registered trademark of the IBCS Institute. Fennibi is based on the IBCS® notation rules, but is not IBCS®-certified and is not affiliated with the IBCS Institute or the IBCS Association.

Rebuild Your Reporting — in Days, Not Months Preview

Power BI report landscapes that have grown over the years are fragmented: every report is built differently, and readers struggle to find their way around. Instead of painstakingly reworking each report, you start fresh. All you need are the business questions your existing reports are meant to answer.

Collect the questions

Which questions should your reports answer? This list is the real business substance of your reporting.

Ask them in the chat

Fennibi answers each question straight from your Power BI model, with a chart, a table and a conclusion.

Save them as reports

Good answers become reports that are up to date every time they are opened. The reporting team maintains them centrally, and anyone can add their own.

Every report looks the same: one consistent presentation based on IBCS® and a PDF at the click of a button. Forecasts for simple cases are included too, without having to call on the modelling team.

The Foundation: Documentation the Assistant Understands

An assistant is only as good as the model it reads. Fennibi documents every Semantic Model fully automatically — tables, measures, DAX formulas, import logic, security — and turns that into your searchable data catalog. The very documentation your colleagues read is what every live answer is built on.

Browse the example data catalog →

What You Gain with Fennibi

Live analysis

Ask your own Semantic Model, calculated live in your Fabric capacity — with charts, tables, and a conclusion backed by the figures.

Unified reporting

Reporting built from your business questions: set up in days, every report structured the same way, all in one place.

Results you can read

Charts and tables in one consistent notation based on IBCS® — comparable across every analysis you run.

Automated data catalog

Fennibi extracts the metadata of all Semantic Models fully automatically and keeps the catalog up to date — with no manual effort.

Model diagnostics

Detects missing descriptions, unused columns, and modelling anti-patterns automatically.

Version changelog

Automated diff between two model versions: added, changed, removed — traceable at a glance.

Audit-proof PDF/A

Immutable, dated PDF/A documentation as audit-proof evidence — for MaRisk, GoBD, and regulatory reporting.

Spot consolidation

Surfaces overlapping and redundant models — the basis for streamlining your model landscape.

Immutable PDF Documentation of Power BI Models

For the reporting of banks and financial institutions, MaRisk (e.g. AT 4.3.4) requires traceable, documented data management. Fennibi turns every Power BI Semantic Model into an immutable PDF/A document — dated, versioned, and audit-proof.

  • ✓ Immutable archival PDF (PDF/A-2b) as audit-proof evidence of your data models
  • ✓ Every document carries its data cut-off date, version number, and generation timestamp — auditable for regulators
  • ✓ Complete model documentation for regulatory reporting (MaRisk, GoBD)
  • ✓ Consistent, automatically generated evidence across all Semantic Models

Prepare Your Power BI Models for the Fabric Migration

Before migrating to Microsoft Fabric you need a complete overview of your existing Power BI landscape. Fennibi delivers the clean inventory your migration builds on.

  • ✓ Complete catalog of all Semantic Models as the starting point for the migration
  • ✓ Identify and consolidate overlapping and redundant models before migrating
  • ✓ Uncover inconsistent KPI definitions before they are carried over into Fabric
  • ✓ Diagnostics find performance bottlenecks in import logic and DAX — so no legacy baggage is carried into Fabric

The name

The fennec fox listens closely with its big ears. Hence the name: bi as in Business Intelligence, and bi as in two, because one fennec listens to the business side, the other to the technology.

See more

Personal Demo & Individual Questions

In a quick 30-minute Teams call I can walk you through Fennibi with your specific use case in mind — in English or German. Just suggest three possible time slots (include your timezone if outside Europe/Paris/Brussels) and I'll do my best to make one of them work.