Features built for an AI-Native enterprise platform
Every feature in Netgrif is designed around one principle: the Petriflow model is the application. AI generates it, humans refine it, the platform interprets it directly. From architecture to runtime, every layer serves that loop.
Modern, Open Architecture
A three-tier stack:
- Backend: Java/Spring Boot
- Frontend: Angular
- Data Layer: MongoDB and Elasticsearch
Open source foundations you can inspect, scale, and extend. No black box, no vendor lock-in.
Petriflow — the AI-Native Modelling Language
One model defines:
- Workflows & tasks
- Data, forms, and actions
- User roles and lifecycle states
Deterministic enough for an LLM to generate reliably. Expressive enough for humans to build with. The model is the running application — not a design artifact, not source code that needs compiling.
Integrated Workflow + UI in one artifact
Process logic, forms, permissions, and UI behaviour live in the same Petriflow file. AI generates them together; the platform runs them together. No separate frontend repo, no schema-versus-code drift.
Transparent,
All-Inclusive Licensing
One commercial license. Everything included.
The full AI-native platform — builder, runtime, and AI assistant — under a single license. No hidden modules, no forced upgrades, no per-seat surprises.
A fully functional Community Edition is also available under a source-available license — explore and build with Netgrif at no cost.
Petriflow itself is open-source. Inspect the language, extend the engine, reuse it in your own environment or research — it belongs to the community.
Deploy wherever your data lives
- On-premise for full sovereignty
- In the cloud for scale and speed
- Hybrid for regulated industries — banks, insurers, healthcare
Direct access to every layer — backend (Java), frontend (Angular), and data (NoSQL). Customize deep when you need to; let the AI handle the rest. Especially suited to regulated and complex domains where off-the-shelf solutions fall short.
Connect to ERP, CRM, IAM/IDM, and bespoke systems using REST, SOAP, and standard messaging protocols. Every Petriflow model can call out — or be called — through open APIs, so AI-generated apps integrate cleanly into your existing IT ecosystem.
NoSQL architecture on MongoDB and Elasticsearch — flexible schemas, high-volume operations, no built-in retention limits. Long-term traceability and historical insight without performance trade-offs, ready for the data volumes that AI-driven processes generate.
A native BPM engine interprets Petriflow directly — workflows, task routing, and UI behaviour all run from the same model. No separate orchestrator, no schema mapping, no glue code. The spec is the runtime.
Every Petriflow model is versioned, every action logged, every role enforced via RBAC. AI-generated output is verifiable — readable by humans, traceable through audit, and editable when needed. Built for compliance-driven environments where you must prove what ran, when, and by whom.
Best fit for
Who builds on Netgrif
- Enterprises modernizing legacy BPM and form-based systems with AI-native tooling
- Teams generating internal portals, self-service apps, or partner workflows from natural language
- Regulated industries — banks, insurers, healthcare — that need AI output to be verifiable
- Hybrid teams where business analysts, developers, and AI collaborate on the same model
- Projects that need rapid AI-driven delivery without sacrificing process transparency