Flexible licensing. Start small. Scale to enterprise.

On Entry and Pro, you provide your own Azure OpenAI API key. Cadenza controls the prompts, temperature, parameters and safety guardrails — but the inference happens on your tenant, so usage is billed by Microsoft to you directly. This keeps customer data outside our infrastructure for AI workflows. Enterprise customers can opt for a fully managed, custom-trained model instead.

30 days of full Pro-tier access on a single project of your choice — no credit card. We'll typically scope it against your real Polarion or DOORS project so the AI is working on data that actually represents your audit risk. Reach out via the contact page to start one.

No. Cadenza reads from your ALM on demand, processes in-memory, writes results back through the same REST API your team already uses. We cache aggressively for performance, but the canonical store is always your ALM — uninstalling Cadenza never destroys data because the data was never ours.

Yes, on Enterprise. We ship Docker images and an Alembic-managed PostgreSQL schema. Many automotive customers run Cadenza fully air-gapped, with a self-hosted LLM (vLLM, Ollama, or a fine-tuned proprietary model). Our team supports the install.

A user is any human seat that authenticates to Cadenza in a billing month. Read-only viewers (e.g. an assessor reviewing a frozen report) and service accounts (CI/CD integrations) are free.

Yes — annual commitments receive a discount. Enterprise contracts are typically multi-year with custom payment terms. Talk to us via the contact page for a quote.

Cadenza generates audit-grade reports (DOCX, PDF) and read-only project snapshots that your assessor can review without needing a Cadenza seat. If your assessor wants to interactively interrogate the link graph, we provide free observer seats.