Power BI Model Health Snapshot
Your reports aren’t the problem. Your semantic model is. Upload or paste your Tabular BIM JSON or TMDL files below to audit star-schema compliance, bidirectional cross-filtering traps, and VertiPaq memory bloat.
Moderate architectural health (84/100). Identified 2 high-impact issue(s) that should be addressed before enterprise scaling.
Evaluates dimension depth, snowflake normalization, and surrogate key hygiene.
Flags bidirectional cross-filtering cascades, M:M bridges, and inactive paths.
Audits timestamp column splits, Auto Date/Time sprawl, and high-cardinality bloat.
Detects unreferenced measure sprawl, iterator loops, and missing VAR blocks.
Checks TMDL source-control readiness and calculated columns in Direct Lake.
Change cross-filtering to single direction (OneDirection). If bidirectional filtering is required for a specific visual or metric, activate it locally using CALCULATE(..., CROSSFILTER('Table'[Col], 'Other'[Col], Both)).
Denormalize attributes from the secondary dimension into 'Dim_Product' in Power Query or upstream data views to restore a clean 1-hop Star Schema.
Want a senior Power BI engineer to audit these findings?
Automated tools find structural symptoms. My fixed-scope Health Check reviews VertiPaq memory, DAX, refresh design, and development workflow by hand, and ends in a prioritized remediation plan.
Privacy: the analysis runs in your browser and your model metadata never leaves it. I only count that the tool was used and when a report is downloaded — no inputs, results or personal data are recorded.
Get the next Engineering Note
Practical DAX, semantic model and Fabric notes. 1–2 a month, unsubscribe any time.
The 3 Pillars of Resilient Semantic Models
High-performing Power BI environments are built on clean physical data architecture. Optimizing DAX syntax cannot overcome a fragile or ambiguous data model.
Star Schema Topology
Single-hop 1:* relationships from dimensions to fact tables enable vectorized VertiPaq scans. Snowflake chains increase formula engine traversal overhead.
Single-Direction Filters
Bidirectional relationships allow filters to flow backward, causing ambiguous query paths, unexpected visual totals, and severe query plan thrashing.
VertiPaq Dictionary Encoding
High-precision timestamps and surrogate primary keys in fact tables prevent dictionary compression, consuming excessive RAM in Power BI capacities.
Model Diagnostics & Privacy Standards
Is my semantic model metadata secure?
Yes. All parsing and rule evaluation occurs 100% client-side inside your browser's memory using local JavaScript. Zero table names, column names, measures, or data are transmitted over the network or stored on my servers.
What files can I upload to the Model Health Snapshot?
You can upload or paste a model.bim file (from Tabular Editor or Visual Studio), a TMSL script, a TMDL text definition (from Power BI PBIP project folders), or sanitized JSON schemas.
Why are bidirectional relationships flagged as high risk?
Bidirectional relationships allow dimension tables to be filtered by fact tables and vice-versa. In complex models with multiple fact tables, this introduces ambiguity, circular dependency risks, and forces the Formula Engine to materialize large intermediate tables rather than leveraging multi-threaded VertiPaq storage engine compression.
How does this free tool differ from the Power BI Engineering Health Check?
The Model Health Snapshot is a static structural audit of table schemas, relationships, and metadata. The paid Power BI Health Check goes further: I review your model metadata and VertiPaq Analyzer export by hand, check DAX efficiency and refresh query folding, and deliver a written report with a prioritized remediation plan.
Go from static metadata to VertiPaq engine metrics
In my fixed-scope Health Check I review your model metadata by hand: column memory from a VertiPaq Analyzer export, relationship design, DAX efficiency, and refresh query folding.