QorVertex is an AI-enabled system of intelligence for your source code. It maps every layer of your software, from the UI through services, APIs and events down to the database, across every repository in your enterprise, in 8 languages. Then you ask it anything, in plain English, in its own chat or inside Claude Code, Cursor and Copilot.
A coding assistant sees the repository you have open. QorVertex’s AI works from the map of every repository, so it can answer the questions that cross them, and it shows the file and line behind every claim. See how it works →
Start from whatever is on your screen, a class, a file, a message channel or a blast radius, and ask about it. Every answer shows the queries it ran, and exports as a report. The model runs in your cloud account, on AWS Bedrock or Google Vertex AI; nothing is sent to us.
Your cloud, your governanceAn MCP server gives Claude Code, Cursor and GitHub Copilot the same map, so they can answer about repositories they have never opened, citing file and line. No model provider needed: your assistant brings its own.
9 read-only toolsThis is Article, a class in the open-source Spring Boot RealWorld app, in our live demo. Ask what a change to it would touch, and QorVertex lists 201 places within two hops: the classes that import it, the services and data-access code that use those, and the tests that exercise them, each with its file and line.
The answer came back in under half a second. On your own code the same walk keeps going across repositories, APIs, message channels and shared tables.
A change to one shared orders collection reaches five services in five repositories, and nothing in your IDE will warn you, because those services never reference each other in code. They only share a name.
QorVertex scans every repo into one catalog and treats shared tables, stored procedures, and message queues as first-class dependencies. So the coupling becomes something you can see, search, and measure.
QorVertex scans every repository and works out what connects them: the calls, the APIs, the message channels and the shared tables. People explore that map directly; AI answers from it, so it can cite where every answer came from. Explore the product →
Pull every repository across GitHub, GitLab, Bitbucket, and Azure DevOps into one catalog. Auto-rescan on check-in — no public webhook required.
4 VCS providersCalls, imports, inheritance, and type bindings resolved across project and repository boundaries — with durable edges that survive partial rescans.
The coreShared tables, stored procedures, and physical databases as first-class nodes — with per-column blast radius and cross-language coupling.
Shared DataWho exposes and who consumes each endpoint — REST, gRPC, GraphQL, MCP, and SOAP — with "Called by N" across every repo.
5 protocolsPublishers and subscribers linked through channel names across 10 broker families — Kafka, RabbitMQ, SNS/SQS, Pub/Sub, Sidekiq, JMS, and more.
10 broker familiesWalk dependents transitively from any symbol, table, column, endpoint, or event. Answer "what breaks if I change this?" before you commit.
Change safetyUML class and component diagrams, container diagrams, React component trees, dependency graphs and shared-table data-flow views, all drawn from the code, with full version history, diff, and restore. See one →
DiagramsFind any class, function, table, endpoint, or package — and every place it's used across all repos, ranked by real usage.
Find anythingAn MCP server that lets Claude Code, Cursor, and Copilot answer about repositories they have never opened — with file:line. No model provider needed; your assistant brings its own.
9 read-only toolsWhich integration and end-to-end suites must run for this change — across repositories. Always a minimum, never an allow-list: what can't be analysed says run everything.
Run less, safelyAn industry XSD, WSDL, or proto compiled into many products couples them by name alone. Trace a schema type through the generated class to every line of code that touches it.
XSD · WSDL · ProtobufWhat a component is deployed into, and what it talks to at run time — Docker Compose, Kubernetes and Helm, JBoss, WebLogic, IIS — wired into the same blast radius.
Compose · K8s · app serversThe blast radius of a single column change, walked across services and back to the code that calls it. Explore it live → — no signup, a curated public catalog.
From source and APIs to databases, message brokers, and how it all deploys — coverage is what makes an org-wide graph possible. See the full list →
Search any symbol, table, endpoint, or package and see every consumer, grouped by repository — not just the ones in the file you're looking at.
→ Org-wide search & dependentsWalk the blast radius transitively — through shared columns, procedures, and message channels — before the change ships.
→ Impact analysisEvery producer and consumer of a shared node, across repositories and languages, on one page with read/write breakdown.
→ Shared DataAdd your source connections. QorVertex clones and parses every repository — TypeScript, JavaScript, Python, Java, C#, Go, PHP, and Ruby, tree-sitter accurate — into a central PostgreSQL catalog.
Symbols, APIs, events, and database access are resolved into shared nodes, so a table or endpoint many services touch becomes one thing they all point at.
Ask in plain English, search, walk the graph, and measure the blast radius of any change. Auto-rescans keep the map current as code lands.
One downloadable bundle, one install.sh. It provisions the whole stack behind an nginx ingress on a single Linux box — your source and your catalog never leave your network.
Self-hosted at every tier. Start free, grow by lines of code analyzed.
Prices are placeholders — tiers are being finalized. Metering is by lines of code scanned. During the early adopter program, selected organizations get a 90-day evaluation; the free tier opens at general availability.
Apply to the early adopter program and we'll help you install QorVertex on your own code. See your organization's real dependency graph — database included.
