Frequently Asked Questions

Every organisation is a fully isolated tenant with role-based access control, granular permissions, and a complete audit log of every action. Sensitive credentials are encrypted at rest and never shown again after saving. Documents support per-role access, passwords, expiry dates, and access-count limits.

Yes. Documents and folders carry per-user and per-role permissions with view, edit, and full-control levels, plus optional passwords and expiry dates. Workflow stages are owned by named users or workgroups, and only the designated final approver can record the outcome.

A workflow is a chain of task stages, each with an owner. Submitting a connected form starts the workflow and routes the first step to its owner. Owners move work forward or back, refer it, or pause it, and decision stages branch on approve, deny, query, or recommend. Every completed run lands on a completion report with the final outcome.

Yes. Test Run routes every stage of the workflow to you, so you can rehearse the entire process end to end without involving the real task owners. When you are satisfied, activate Runtime and the workflow goes live across your organisation.

Yes. You can sign in with Google, Microsoft, or GitHub, use passwordless magic links, and protect accounts with email OTP or an authenticator app. Policies cascade from platform to organisation to user, so admins can require 2FA for everyone.

The Workspace Agent answers questions about your workspace, navigates you to resources, generates forms from a prompt, and guides workflow construction. It can also perform actions like submitting or approving, but only after your explicit confirmation and only within what your role allows.

Yes. Every organisation ships with a Consultant role and an Internal Reviewer role. Both open Onboarding and Organogram and nothing else, so an external consultant can work a dossier end to end without ever seeing personnel records, leave, attendance, or payroll. An admin can grant either role alongside whatever role someone already holds.

A consultant proposes a change rather than overwriting the client's answer, so the old and new values sit side by side. The nominated internal reviewer is notified and can accept, modify, or reject it. Every proposal and decision is attributed with a name and timestamp, and the dossier ends with a recorded consultant sign-off and client sign-off.

The dossier captures departments, units, processes, and role responsibilities from the Org Onboarding Template. From there you can build your live organogram, attach role profiles to positions, generate workflows from the captured processes, and export the whole thing as a Standard Operating Manual in PDF or Word.

Yes. Organisations can bring their own database, S3-compatible storage, vector store, and LLM provider, including OpenAI, Anthropic, Google AI, Azure OpenAI, OpenRouter, and Ollama. Usage is metered per provider and model so you always know what is being consumed.

Yes. The platform provides the controls NDPR expects you to have: granular access control over personal records, an audit log of every read and write, and the option to keep data in your own database and storage in a country you choose. Your organisation remains the data controller, and Approovia provides the tooling and evidence trail to demonstrate accountability.

You choose. An organisation admin can point the platform at your own PostgreSQL or MongoDB database and any S3-compatible bucket, hosted in the region or data centre you require. Connection details are encrypted and never shown again once saved, and files never touch platform storage once your own bucket is configured.

Yes. The platform ships as containers, so it runs on AWS, Azure, Google Cloud, a private cloud, or hardware in your own building. With your own database, storage, and a self-hosted model, an installation can operate with no outbound internet access, which suits government and enterprise workloads with strict isolation requirements.

Traffic runs over TLS between the browser, the gateway, and internal services, and the browser only ever talks to a server-side proxy so tokens are never exposed to page scripts. Sensitive configuration such as connection strings and provider keys is encrypted at rest with AES-256-GCM, and sessions live in an HTTP-only cookie verified on every request.

You can bring your own. Chat and embedding models are configured separately, and supported providers include OpenAI, Anthropic, Google AI, Azure OpenAI, OpenRouter, and self-hosted Ollama. Because the key is yours, AI spend sits on your own account at your own rate, and usage is metered per provider and model with platform calls shown separately from your own.

Work can sync with external trackers like GitHub, JIRA, Linear, and Plane.so, with identity mapping so assignees resolve correctly. API tokens let you integrate your own systems, and lists export to CSV, Excel, and PDF.