Yantra Studio — frequently asked questions

What is Yantra Studio used for?

Yantra Studio is used to model architecture and keep the model reviewable. You draw solution and enterprise architecture, model data structures, record how components depend on each other, and attach the governance content — principles, risks, non-functional requirements — to the shapes themselves.

It is aimed at work that outlives a single meeting. Diagrams can be versioned, put through an approval workflow, and stored in GitHub, so the picture and the decision behind it stay together.

Diagram ▸ Approval workflow · Diagram ▸ Version history · Arch ▸ GitHub

Do I need to be an enterprise architect to use Yantra Studio?

No. The drawing surface works like any diagramming tool: you insert shapes, connect them, and lay them out.

The architecture-specific parts — viewpoints, principle compliance, tech debt reporting — are separate menus you can ignore until you need them. A developer documenting one service and an enterprise architect maintaining a portfolio use the same canvas.

Page ▸ Entity / Table · Page ▸ Connector · Arch ▸ Viewpoints

How do I start a diagram?

Open a diagram and use the Page menu to insert what you need: an entity or table, a note, a text box, a connector, or freehand drawing. If you would rather not start from an empty canvas, the template gallery is in the same menu.

Shapes beyond the basics come from shape packs, also in the Page menu.

Page ▸ New page · Page ▸ Entity / Table · Page ▸ Template gallery · Page ▸ Shape packs

Why is a menu item greyed out?

Because that capability is not enabled for you. Access is resolved in three levels: a super admin enables a feature area for your organisation or your individual account, then an organisation administrator or you enable specific features inside it.

A greyed item is therefore an administrative switch, not a missing feature. Ask whoever administers your organisation to enable it.

Can I use Yantra Studio to draw a solution architecture?

Yes. Insert components, connect them, and arrange them with the layout tools. Auto-layout offers hierarchical, circular, force-directed and orthogonal grid arrangements, so a diagram assembled in any order can be tidied in one step.

For a system design that will be discussed rather than built immediately, the simulation architecture templates give you a working starting point — a three-tier cloud architecture, an ETL pipeline, and a range of well-known system designs.

Shape ▸ Auto-layout · Shape ▸ Hierarchical (default) · Shape ▸ Force-directed · 3-Tier Cloud Architecture · ETL / Data Pipeline

Can I show how applications connect?

Yes. Connectors between components are the basic mechanism, and each connector is a modelled relationship rather than a drawn line, which is what makes dependency and impact views possible later.

For a portfolio view of platforms rather than individual shapes, Platform Lens presents the diagram through the platforms its components belong to.

Page ▸ Connector · Arch ▸ Platform Lens

Can I model an architecture before it is implemented?

Yes, and this is one of the clearer uses. Model the proposed design, then run it through the analysis tools rather than waiting for production to answer the question.

Simulation covers workload profiles, sensitivity analysis, bottleneck detection and a resilience score. These are model-based estimates from the architecture you have described, not measurements of a running system, and should be read that way.

Simulate ▸ Workload Profile · Simulate ▸ Sensitivity Analysis · Simulate ▸ Bottleneck Detection · Simulate ▸ Resilience Score

Can I prepare an architecture for a stakeholder discussion?

Yes. Presentation mode has two forms: Flow, which walks the diagram in a defined order, and Designer, which follows a script you write.

For a document rather than a live walkthrough, export to PDF, PNG or SVG, or assemble the material through the Export Center.

Page ▸ Present · Page ▸ Flow (default) · Page ▸ Designer (script) · Diagram ▸ Export Center · Diagram ▸ PDF

Can several people work on the same architecture?

Yes. A diagram can be shared by link and has a team and collaborators list. Review is handled by an approval workflow rather than by comments alone, so a diagram can be formally signed off.

Sharing, collaboration and approvals are each separately enabled, so what your organisation has switched on determines which of these you see.

Diagram ▸ Share link · Diagram ▸ Team & collaborators · Diagram ▸ Approval workflow

Can I model a database as well as an architecture?

Yes. Entities and tables are first-class shapes, and the quick-create templates start you at the right level of detail: a conceptual ER model, a logical model, or a physical model.

Columns, keys and relationships are modelled rather than drawn, which is what lets the same model produce SQL later.

Page ▸ Entity / Table · Star Schema · Event Sourcing

Can I generate SQL from a model?

Yes. DDL can be exported for PostgreSQL, MySQL, BigQuery and SQL Server.

For a model that already exists in a database, Delta DDL produces the migration between what you have modelled and what is deployed, rather than a full create script.

Diagram ▸ DDL (SQL) · Diagram ▸ PostgreSQL · Diagram ▸ MySQL · Diagram ▸ BigQuery · Diagram ▸ SQL Server · Data ▸ Delta DDL / Migration

Can I keep business meaning alongside the technical model?

Yes. A data dictionary and a business glossary are maintained with the model, and subject areas group a large model into readable parts.

This matters when the same table means different things to different teams: the definition lives with the model rather than in a separate document that drifts.

Data ▸ Data Dictionary · Data ▸ Business Glossary · Data ▸ Subject Areas · Data ▸ Metadata Repository

Can I check a model against standards?

Yes. Naming Standards checks names against the conventions your organisation has defined, Validate Model checks the model itself, and Health Report summarises the result.

These are checks against declared rules. They tell you where the model departs from the standard; they do not rewrite it for you.

Data ▸ Naming Standards · Data ▸ Validate Model · Data ▸ Health Report

Can I edit many objects at once?

Yes. The Bulk Editor changes properties across many objects in one pass, which is the practical way to apply a naming change or a classification across a large model.

Data ▸ Bulk Editor

Can I use Yantra Studio to understand dependencies?

Yes. Because connectors are modelled relationships, the dependencies between components can be read rather than inferred by eye. Lineage traces how data moves through the model.

This is the difference between a picture of a system and a model of it: the same drawing answers "what feeds this?" without you tracing lines manually.

Data ▸ Lineage

Can I see what a change would affect?

Yes. Impact Analysis reports what depends on the thing you are proposing to change, following the relationships recorded in the model.

Its answer is only as complete as the model. Dependencies that were never modelled cannot appear in it, so treat the result as a floor rather than a guarantee.

Data ▸ Impact Analysis

Can I map a source system to a target system?

Yes. Source→Target Map records field-level mapping between systems, and Data Manager is where the wider data landscape is maintained.

This is the usual starting point for a migration: state the mapping first, then use impact analysis to see what the migration disturbs.

Data ▸ Source→Target Map · Data ▸ Data Manager

Can I compare environments?

Yes. Env Compare shows the differences between environments, which is the fastest way to answer why something behaves differently in production than in test.

Data ▸ Env Compare

Does Yantra Studio come with architecture templates?

Yes. There are data-modelling templates — Star Schema, SaaS Multi-tenancy, an Amazon-style e-commerce platform and Event Sourcing — and quick-create ER, logical and physical model starts.

Alongside those are 28 simulation architecture templates covering common system designs, from a three-tier cloud architecture and an ETL pipeline to worked examples such as a URL shortener, a rate limiter and a distributed message queue.

Page ▸ Template gallery · Star Schema · SaaS Multi-tenancy · Amazon-style E-commerce Platform · Event Sourcing · URL Shortener (TinyURL) · API Rate Limiter · Distributed Message Queue (Kafka)

Can my organisation add its own templates and patterns?

Yes. Organisations maintain their own pattern library and template set through the administration screens, so an approved reference architecture can be offered to everyone rather than copied by hand.

Can I import an existing database schema?

Yes, three ways: import a DDL SQL script, import structure from CSV or Excel, or connect to a live database and import directly.

Live database import and DDL import are separately enabled, so which of these you have depends on your organisation.

Data ▸ Import DDL / Mermaid ER · Data ▸ Import CSV / Excel · Data ▸ Live DB Import

Can I bring diagrams in from another tool?

Lucidchart files can be imported directly as .vsdx. Other diagrams can be brought in through Open Format File.

An import produces a model you then work with, not a picture. Expect to review how shapes were interpreted rather than assuming a perfect translation.

Diagram ▸ Import Lucidchart (.vsdx) · Diagram ▸ Open Format File

Can I import an architecture generated from source code?

Yes, through the Archify integration. Archify analyses a repository and writes an architecture JSON file; importing it creates the components, connections and boundaries on the page.

Importing the same document twice merges rather than duplicates: shapes you moved stay where you put them, and anything the code no longer contains is reported rather than deleted. The export goes the other way, so the model is not trapped here.

Diagram ▸ Import Archify (.json) · Diagram ▸ Architecture model · Diagram ▸ Archify (.json)

Can imported components link back to the code they came from?

Yes, when the imported document carries that information. An Archify document whose components declare sources is imported as evidence, and the Trace tab of the Properties panel shows the file and line range behind each shape.

A link that opens the code at the pinned commit is offered only when the evidence carries a repository and a full commit SHA and the repository is on GitHub or GitLab. Where it does not, no link is shown rather than a guessed one.

Can I merge two diagrams?

Yes. Merge diagram combines another diagram into the current one, which is the usual way to assemble a portfolio view from models maintained separately.

Diagram ▸ Merge diagram

Can I document risks against an architecture?

Yes. Risks are recorded against the architecture and can be displayed as an overlay on the diagram, so a risk is visible on the picture rather than filed away in a register nobody opens.

There is also an assisted route that proposes risks from the diagram for you to accept or reject.

Arch ▸ Overlays · Arch ▸ Suggest risks from diagram

Can I document non-functional requirements?

Yes. NFRs are held against the model, and the Tech Debt Report reads the same content to show where the architecture is carrying debt.

Recording an NFR against the component it constrains is what makes it reviewable later, rather than leaving it in a requirements document nobody consults during design.

Arch ▸ Tech Debt Report

Can I check a design against architecture principles?

Yes. Principle Compliance reports the model against the principles your organisation has defined, and Policy & Standards covers the wider rule set.

These report compliance; they do not block work. The point is to make a departure from a principle visible and deliberate.

Simulate ▸ Principle Compliance · Simulate ▸ Policy & Standards · Simulate ▸ Rule Evaluation

What are viewpoints for?

A viewpoint presents the same model for a particular audience or concern, so a security review and a delivery plan can read one model rather than maintaining two diagrams that disagree.

Arch ▸ Viewpoints

Can architecture work be versioned and approved?

Yes. Diagrams have version history and branches, an approval workflow for sign-off, and can be stored in GitHub so architecture changes sit in the same review process as code.

Versioning, approvals and GitHub storage are separately enabled.

Diagram ▸ Version history · Diagram ▸ Branches · Diagram ▸ Approval workflow · Arch ▸ GitHub · Arch ▸ Repository

Is there an audit trail?

Yes. An audit log records activity, which is what makes a governed model defensible when someone asks who changed what.

What does the AI assistance actually do?

There are four assistants: a model assistant for the data model, a SQL analyser, an OpenAPI specification reader, and an architecture copilot. Simulation has its own copilot.

They propose; you decide. Output is presented for review rather than applied to the model on your behalf.

AI ▸ AI Model Assistant · AI ▸ SQL Analyzer · AI ▸ OpenAPI Spec · AI ▸ Architecture Copilot · Simulate ▸ Simulation Copilot

Is simulation a measurement of my real system?

No. Simulation reasons about the architecture you have modelled. Bottlenecks, resilience scores and sensitivity results are estimates derived from that model and the workload profile you supply.

They are useful for comparing designs and finding obvious weaknesses before building. They are not a substitute for load testing the real system.

Simulate ▸ Simulation Panel · Simulate ▸ Workload Profile

How complete is impact analysis?

It is exactly as complete as the model. Impact analysis follows recorded relationships, so a dependency nobody modelled is a dependency it cannot report.

Use it to find what you would have missed, not to prove that nothing else is affected.

Data ▸ Impact Analysis

Will every capability be available to me?

Probably not, and that is by design. Most capabilities sit behind a feature switch resolved per organisation and per user, so two people in the same company can see different menus.

If something described here is greyed out, it is an administrative setting rather than a missing feature.

Does importing from another tool preserve everything?

No import preserves everything. An import maps another tool’s concepts onto this model, and anything with no equivalent cannot survive the trip.

The Archify import is explicit about this: it reports what it was unsure of and lets you correct it before committing, rather than guessing quietly.

Diagram ▸ Import Archify (.json)