Documentation Menu

Frequently Asked Questions (FAQ)

Short answers to the recurring questions that come up in sales conversations, technical discovery, and investor presentations.

01Setup and Deployment

Can BriqMind work without a cloud requirement?

Yes. BriqMind is positioned to support on-premise, private network, and air-gapped deployment scenarios. It can align with your organization's security standards.

Is our data used for model training?

No. The core commitment in the documentation is that customer data is not used to train foundation models. Your data is processed in an isolated environment.

How quickly can a POC turn into a production decision?

It depends on scope, but a pilot with clear success metrics is usually completed in 2-4 weeks and converted into a decision for enterprise rollout.

Which teams see value fastest?

Software, operations, finance, legal, and knowledge-heavy teams usually see the first concrete gains. Any department with repeatable processes has potential.

Which documents are critical for procurement teams?

Trust Center, Security, SLA & Support, On-Premise, and Production Readiness are the core package for procurement and risk review processes.

02Security and Data

How do Birk models manage hallucination risk?

Birk models use Retrieval-Augmented Generation (RAG), grounding responses in your organization's approved data sources and producing answers with references.

Can it integrate with our existing authorization infrastructure such as Active Directory or SSO?

Yes. SAML, OAuth2, and OIDC standards are supported. Birk can integrate directly with Active Directory, Entra ID, or Okta while preserving data access permissions.

Which data sources can we connect directly?

Ready connectors are available for Confluence, Jira, SharePoint, Google Drive, OneDrive, GitHub, GitLab, and common relational databases.

Do you provide API access?

Yes. All processes are designed API-first. You can build internal apps, automations, or custom interfaces on top of the Birk API.

03Operations and Support

How does your pricing model work?

We offer three main models. First: monthly per-user licensing for SaaS, with entry packages around USD 49 per user/month and standard API usage quotas. Second: consumption-based API pricing, starting around USD 0.40 per 1M input tokens for Birk-Fast and around USD 9 per 1M input tokens for Birk-Heavy; output tokens are priced at roughly twice the input rate. Third: annual enterprise/on-premise licensing, starting from a USD 75,000 annual base depending on hardware capacity, concurrent users, and support level. Air-gapped deployment, industry-specific models, and SLA packages are added as separate modules. These figures are indicative; final pricing depends on scope, SLA, and support level.

Which languages are supported?

Multilingual support is available, with Turkish and English as primary languages. Birk is especially optimized for Turkish understanding and generation.

How are system updates handled?

In the SaaS model, updates are applied automatically. For on-premise deployments, zero-downtime update packages are provided periodically.

Is logging and audit infrastructure available?

Yes. Every query, accessed document, and system event is logged in detail. These logs can be forwarded to your SIEM tools.

How long does deployment take on average?

With SaaS, you can usually start the same day. On-premise or private network deployments typically take 1-2 weeks depending on infrastructure readiness.

04Usage and Expansion

What can we do to improve output quality?

Output quality improves directly with clean connected data sources, structural consistency, and good prompt engineering.

Is mobile access possible?

Yes. In addition to the responsive web interface, native mobile apps will soon provide access from anywhere.

What is the orchestration layer?

The orchestration layer allows multiple agents and models to work together, split tasks among themselves, and manage more complex processes.

Can we use our own models with Bring Your Own Model?

Enterprise plans can support integrating your own trained open-source models or custom fine-tuned models in compatible environments.

Are there user limits?

This depends on your license model. Enterprise plans can support unlimited users as far as your hardware or cloud quota allows.

Do you offer SLA-backed support when needed?

Yes. Enterprise customers can receive 24/7 critical incident support, a dedicated customer success manager, and SLA packages with 99.9% uptime commitment.