Security & Governance

Built to be
trusted and governed.

Before a solution goes live, technology teams need to know who can access what, where data goes, and how every agent stays under control. Here's how Qlar answers.

Access control

Decide who can talk to each agent and what each person may see.

  • Private or public agents, login required
  • Knowledge filtered per user
  • Row-level security in databases

Roles & audit

Clear roles across organizations, and a record of who changed what.

  • Organizations and roles (Owner, Admin, Builder, Billing)
  • Per-agent contributor rights; every access to history or personal data logged
  • Audit trail from old value to new

Sensitive documents

Stop sensitive files before they enter a conversation.

  • An upload policy written in plain language
  • Every image and document checked at upload, in chat and voice
  • Violations refused with a reason

Internet access

Decide whether, and where, agents may reach the web.

  • Web search on or off
  • Allowed domains only
  • Web scraping on or off, with excluded paths

Privacy & retention

Keep nothing longer than your policy allows.

  • Privacy Mode encrypts conversations
  • Deleted within 1–30 days, unrecoverable even by Qlar
  • No memory, query logs, or analysis while it's on

Models & keys

Choose the AI behind each agent, on your terms.

  • Built-in model levels or your own provider keys
  • Your keys encrypted (AES-256)
  • Admins choose which models are allowed

Cost governance

Keep AI spending predictable as usage grows.

  • Credit limits per agent
  • Budget groups per user
  • Auto top-up with a monthly spending limit

Quality & oversight

Test before every change and keep people in the loop.

  • Automated conversation tests
  • Analytics, logs, and AI conversation evaluation
  • Escalation and human handoff

Automation policies

Automations run inside the limits your organization sets.

  • Organization-wide policy and limits
  • Control over who may create automations
  • Run history and monitoring

Questions from your security team?
We'll walk you through it.

We're happy to walk your technology and security teams through architecture, data handling, and deployment options.

Talk to our teamRead the documentation