Documentation for software architects — created with AI: dev-doc v3.0

Dev-Doc reads a repository — modern or decades old — and produces the overview, the architecture, the module map and the interactive C4 diagrams your organisation was supposed to have, derived from the code rather than from memory.

Get 40 Free Credits See demo projects Executive briefing

40 free credits on sign-up · no credit card · sign in with GitHub, Google or email

Dev-Doc turns source code into documentation

Dev-Doc is a comprehensive solution that covers all the documentation components of a software project. Built on the latest technologies, it meets the needs of developers and stakeholders alike, and provides a standardised approach for medium and large software projects.

Because the documentation is generated by AI from your source code, it is produced fully automatically. Your senior developers are spared the tedious work of writing it, and new team members save a lot of time learning the system architecture.

Core features

One solution for all documentation

One place for all documentation

Ready to go in 5 minutes

Fully automated code docs with AI

Auto-updates on each release

Fully auto-generated visual software architecture

How it works

Three steps from a repository to living documentation

1
Import your project
Paste a GitHub, GitLab or Bitbucket link — private repositories with a read-only token — or upload a zip file. The analysis runs in the background.
2
Dev-Doc explores it
Modules, languages, frameworks, classes and call relationships are read from the code first; then the documentation is written from those facts.
3
Read, review, share
Open Overview, Architecture, Governance and Codebase. Edit and approve sections, share a read-only link, or download Markdown and arc42.

Not ready to import? Explore fully analysed sample projects — architecture diagrams, C4 models, decision records and annotated code. Read-only, and no account needed.

Inside an analysed project

Four views answer four questions

Every project opens with the same tabs. Codebase is optional — you decide at import whether to include it.

Where?

Overview

The big picture first — what the project does, the modules it is built from, its API surface, the tech stack behind it, and how far the documentation has progressed.

What?

Architecture

A full C4 model in four levels — context, containers, components and code — grounded in the real call graph, plus a written architecture description.

Why?

Governance

Inception and communication canvases, architecture decision records and configuration records: vision, key decisions, risks, alternatives and consequences.

How?

Codebase

Optional semantic analysis: every file, class and function explained in plain language — execution flow, key considerations and possible risk points.

Take the full product tour

Communication canvas — Canvas for stakeholders

Stakeholder summary in a canvas

The canvas holds the essential information about the software project and is an important source of information for stakeholders, presented in a particularly clear way.

It deliberately leaves out deep technical detail and focuses on the basics and background. The Inception Canvas and the Communication Canvas follow the arc42 approach — and both can be edited and approved by your team.

Visual architecture

Visual architecture documentation is the centrepiece of Dev-Doc. By analysing every method, we draw your software as it really works — C4 context, container, component and code diagrams, with relationships grounded in the actual call graph and interfaces marked as contracts.

No more hand-made slide decks or extra code to produce diagrams. Oversized diagrams open as a navigable, grouped map instead of an unreadable wall.

C4 context diagram — Architecture for complex projects

Method-by-method documentation — Detailed docs for the dev team

Modules & deployment

The deep technical details of the project in one place, so the documentation becomes a useful resource for the development team: modules with their paths, types, frameworks and entry points, the API surface, the data model, deployment services and the tech stack read from the project’s own files.

Source code explanation

Dev-Doc explains your source code automatically: what each file and method means and how it works, how methods communicate with each other, and a short summary of each file as a whole — side by side with the code.

This helps new project members find their way around faster and cuts recurring questions to senior developers.

AI code explanation — Helpful AI-generated code docs

What sets it apart

Correctness over volume of text

Most of the work goes into proving a statement before it is published.

Legacy estates are first-class

COBOL, PL/SQL, T-SQL and decades-old Java or C are documented with the same depth as a modern web stack.

Facts first, prose second

Modules, deliverables, entry points, dependencies and call relationships are read from code and build files before any text is generated.

An audit after every run

A deterministic check compares each document against the source tree — invented identifiers, missing files, wrong signatures — and records what changed.

Human approval survives

Sections a reviewer edited or approved are kept when the code changes; the rest is refreshed.

Only what actually ships

Tests, examples and benchmarks are documented as non-deployed, so the inventory describes the product you deliver.

Read-only by design

Nothing is installed alongside your code and nothing is written back to the repository. Share links never expose source code.

Working in defence or critical infrastructure?

Engineering evidence package

Software that has been shipping for years, and a customer now asking for the documentation package? From the same analysis, Dev-Doc produces a software design description, configuration-item records with change history, a requirements-to-code traceability matrix, a verification & validation summary — and an SBOM for the Cyber Resilience Act and NIS2.

Compliance & evidence