Security by design, verified at compile time
NPL enforces protocol correctness and permission consistency through compilation, reducing security drift and logic ambiguity.
Build enterprise applications with AI
From business requirements to secure, running software. Build production-ready NPL applications faster with AI while keeping architecture deterministic, auditable, and secure.
NPL gives you compiler-verified protocol logic. AI accelerates delivery. Together, you get speed without losing control.
Security posture
Compiler-verified
Execution model
Deterministic
Assistant support
Tool-agnostic
Why teams choose this
Deterministic, battle-proven stack with no surprises
The workflow follows a stable, proven architecture and strict execution order, producing repeatable outcomes across teams and projects.
Faster delivery from idea to implementation
Users provide basic business requirements, and the agent iterates them into implementation-ready artifacts.
Assistant flexibility
This approach works with Cursor, Claude, and other capable AI-assisted IDE workflows; it is not tied to a single tool.
How it works
Step 0
Start from the official starter
Create your project folder and checkout the starter repository. Then open it in the AI-assisted IDE of your choice.
Copy starter repository command:
git clone https://github.com/NoumenaDigital/noumena-ai-starterStep 1
Enter basic business requirements
Tell the agent to build your app by following strictly the README.md in the project folder. Describe your domain, key roles, core actions, and rules in plain language.
Step 2
Agent structures requirements and runs the workflow
The agent transforms your input into a structured requirements document and then executes the workflow: protocol generation, API generation, identity/access provisioning, and application wiring.
Step 3
Test & iterate
Ask your agent for credentials if they are not already provided, then test your application in the browser. If needed, ask the agent to generate seed data to improve the experience.
Troubleshooting
First action always: give the logs to the agent. Most issues can be resolved fastest when the agent can inspect real command output and service logs.
What you get
By the end of the workflow, you have:
Dockerized system
Local development and execution out of the box.
Keycloak integrated with provisioning
Realm, client, roles, and users setup included.
Fully functioning web application
Production-style frontend with landing page, authentication, and secured user area.
End-to-end implementation of core business process
Your key requirements are translated into executable protocol logic and connected application flows.
Generated API integration
Connected to real backend protocol logic.
Non-functional capabilities
Included out of the box (NPL-centric):
Deterministic protocol execution
business processes run predictably from explicit protocol/state definitions.
Strong transactionality
protocol actions execute with transactional guarantees for state consistency.
Built-in auditability
protocol instance history and action traces support compliance and operational transparency.
Policy-enforced authorization
backend rules and protocol permissions govern access, not only UI checks.
End-to-end traceability
from requirement to protocol to runtime behavior in a single coherent model.
Reproducible environments
containerized infrastructure and scripted setup for consistent results across machines.
Who it is for
Role-based application teams
Teams building role-based business applications.
Engineering organizations
Organizations that want AI speed without unpredictability.
Regulated or auditable domains
Domains requiring deterministic and traceable execution.
Build faster without losing control
Use your preferred AI assistant and ship secure NPL applications on a deterministic, production-oriented stack.