The platform that lets AI think, learn, and act in the physical world.
Twinit provides the software infrastructure to build, run, and manage AI-native digital twin applications. Composable building blocks. Secure runtime. Developer tools that integrate data, knowledge, and AI agents into bespoke solutions. Delivered rapidly.
Most digital twin products are too rigid. Custom development is too slow. Twinit sits between the two: composable infrastructure to build bespoke, AI-native digital twin applications without reinventing the back end.
Twinit is organised in three layers, each with a distinct role in how applications are built, run, and evolved. Together they form the complete development and operations environment for AI-native digital twins.
Twinit Services
The back-end building blocks that developers compose into applications. Each service is purpose-built for the core capabilities digital twin applications need: representing complex physical assets as semantic graphs, orchestrating data flows from IT, OT, BIM, and GIS sources, running AI agents that reason over graph-structured data grounded by retrieval over your documents, event-driven workflows delivering closed-loop actions, enforcing granular access control, and managing application lifecycles without provisioning infrastructure.
Services are composable. Applications are built by combining the blocks you need, not by configuring a rigid product. Every application logic (data models, permissions, pipelines, AI agents, APIs, notifications) is defined in JavaScript scripts that Twinit executes securely and consistently
Semantic Knowledge Graph
Represent physical and virtual systems as living semantic graphs, integrating data from IT, OT/IoT, CAD/BIM, and GIS. AI agents interact with the graph and reason over it. Create events and notifications that drive actions.
Workflows
Orchestrate complex operational processes by capturing state transitions, rules, and dependencies precisely. Schedule recurring logic, trigger on events, and coordinate multi-step workflows.
AI Orchestration
Build and govern teams of AI agents that collaborate to detect anomalies, run simulations, and optimise outcomes. Graph-aware reasoning delivers explainable, trustworthy insights.
​
Integrated Graphics
2D, 3D, process flow diagrams, maps, and game engine integration in a single viewer. BIM model ingestion included. No separate graphics infrastructure required – other than external source systems for maps and game world.
Multi-Source Integration
Native ingestion from OT/IoT via message brokers, BIM via model ingestion tooling, GIS alongside spatial data, and enterprise IT systems via REST and API patterns. Built in File service to manage, view and vectorise files.
RBAC and Authentication
Granular access control at the asset class, location, and capability level. Use Twinit's auth or your own with SSO. Full audit trails for every user action.
​
​
Twinit Workspaces
The secure runtime where digital twin applications run and scale. Workspaces handle data and tenancy controls, secrets and policies, application lifecycle management, and sandbox versus production isolation: insulating applications from infrastructure changes, security risks, and scaling concerns.
​
No provisioning, scaling, or patching of servers. The back end is fully abstracted. Developers build, test, and deploy without leaving the Twinit environment. Applications remain secure, compliant, and resilient as they scale from pilot to production and across customer deployments.
Multi-tenant Isolation
Namespaces isolate tenants via secrets and policies. Sandbox and production environments are separated. Deleting a namespace removes its resources and children ensuring no data leakage between deployments.
​
Audit Trails
Every service emits resource events with identity and timestamp, system and user actions alike: nothing goes unrecorded, and any decision can be reconstructed in full, from the event that prompted it to the human response (Twinit 6.0).
Full Back-end Abstraction
Twinit manages infrastructure, security, and scaling. Applications stay stable and benefit from Twinit Services enhancements over time, without requiring re-engineering each time the underlying infrastructure evolves.
​
Twinit Studio
Twinit Studio is the developer experience for building AI-native digital twin apps. At its core, Twinit's extension for Cursor and VS Code turns natural language into complete Twinit apps, generating idiomatic platform code rather than generic guesses. Around it: a code editor, open APIs and SDKs, reusable templates, and the Twinit Workbench for low-code development. From prompt to production-ready app, in one environment.
Workbench
Twinit Workbench is low-code development through two visual builders: a workflow builder to compose app behaviour, and an ontology builder to define the semantic model the app reasons over
​
​
Open Model Choice
Bring your own model. Twinit's code generation runs on the LLM you choose, with your own keys, so you control which model writes your code, what it costs, and where your prompts and source go. No single-provider lock-in.
IDE Extension (VS Code & Cursor)
Twinit's extension for Cursor and VS Code turns natural language into complete Twinit apps. It pairs frontier models with deep platform knowledge, so prompts generate high quality Twinit code – not generic guesses.