Fed up with high prices and restrictive licensing for your data modelling software?
Try VigilER. A fast desktop modelling tool for data architects with AI review built in, on your own API keys and the model you choose. Conceptual, logical and physical in one file that lives in git. DDL for eight platforms. One simple price, no licence server, no seat audits.
macOS and Windows · model file in git · AI usage billed by your provider at their rates
From whiteboard to warehouse without re-keying anything
Model at the level the conversation is at
Start conceptual: named boxes, definitions, verb phrases, subject areas drawn as regions. Add attributes and keys when the business agrees. Nothing is thrown away between levels.
Let the rules and the AI review it
Deterministic lints catch missing keys, undefined grain, SCD2 without dating, identifiers too long for the platform, names that break your abbreviation standard. The AI reviewer proposes findings with patches you accept or reject as diffs.
Generate, then enforce
Forward engineer a subject area or the whole model to any of eight platforms. Export DDL and dbt files. Export the manifest, and VigilDQ turns keys, foreign keys, domains and SCD flags into checks that run against the data.
Three levels, one truth
Conceptual, logical and physical are the same entities with explicit links. Diagrams are views: a complete model, a subject area, a star, a conceptual overview. Removing something from a view never removes it from the model.
Relational and dimensional together
Facts, dimensions, bridges and outriggers are roles on ordinary entities; surrogate keys, measures and SCD2 dating are roles on attributes. 3NF and star schemas draw from the same model.
AI you control
Paste your own Anthropic or OpenAI key and pick the model. Every finding is validated against the model before you see it; nothing writes your file behind your back. No usage markup.
Your naming standard, enforced
Import the abbreviation list your organisation already keeps. Physical names follow it automatically and a lint flags every table or column that drifted, with the rename as a one-click patch.
Eight platforms
PostgreSQL, Snowflake, Databricks, SQL Server, MySQL, Oracle, BigQuery, Redshift. Dependency-ordered DDL, dbt sources and schema files, and a one-step platform change that re-derives types.
Built for teams and git
One readable JSON file with stable ids: pull requests show real diffs, compare and merge reconciles branches, where-used shows the blast radius of a change, autosave keeps recovery copies.
Simple pricing
Every plan has every feature and every platform. AI usage is billed by your provider at their rates; there is no markup.
Individual
Everything. All platforms. AI review on your own key.
Team
Five seats or more. Same product, volume is the only difference.
Perpetual
A year of updates included. For organisations that cannot buy subscriptions.
Viewer
Open and browse any model, export images. Only the modellers pay.
Support, pay for what you use
| Email support and bug fixes | Included |
| Priority support, next business day | $99 per seat per year |
| Support incident | $150, or five for $600 |
| Onboarding or naming-standards workshop | $1,200 per day |
| Expert model review | $600 per model |
Early access
Builds for macOS and Windows are ready. Leave your email and a line about what you model and I will send a download link and a licence for the early-access period.
Pairs with VigilDQ
The manifest VigilER exports is the input VigilDQ uses to derive uniqueness, referential integrity, allowed-value and current-row checks. Model once, monitor continuously.
About VigilDQ