Unique AI for IT Operators
Unique AI is an enterprise AI platform available as SaaS or self-hosted. It integrates with your data, supports most providers LLMs and embeddings, and provides fine-grained governance across the platform. Every layer — infrastructure, configuration, and integrations — is customisable to your organisation's requirements.
Who This Is For
This documentation is for the technical teams responsible for evaluating, implementing, and operating the Unique AI platform: IT administrators, solution architects, security engineers, infrastructure engineers, and technical project managers.
At a Glance
Section | Description |
|---|---|
Service Models | Deployment options — SaaS and Self-Hosted (Azure, AWS, OpenShift). Architecture overviews, prerequisites, and step-by-step implementation guides for each model. |
Identity and Access Management | Authentication and authorisation architecture, identity federation via SAML and OIDC, user provisioning, and role-based access control across the platform. |
Integrations and Connectors | Data source connectivity through the MCP Hub, available connector types, API integration patterns, and guidance for building custom connectors. |
Operations | Platform update lifecycle, release management, health monitoring, alerting, and incident management procedures. |
Custom Models and Embeddings | Model selection and embedding configuration, including support for bringing your own models and embeddings into the platform. |
How to Use This Documentation
Journey pages follow the implementation lifecycle in order. Each page covers a specific stage with checklists, responsible parties, and criteria that must be met before proceeding.
Technical briefs provide in-depth coverage of a single domain across the full lifecycle. They are referenced from journey pages at relevant points and can be consulted independently.
Playbooks address specific scenarios such as variant deployments, environment promotion, hardening, and incident response. Journey pages link to the applicable playbook when a scenario arises.
Start Here
Follow a process
If you need to... | Start with |
|---|---|
Evaluate the architecture, requirements and security & compliance for a self-hosted environment | Self-Hosted → Evaluate |
Begin a self-hosted implementation | Self-Hosted → Foundation |
Provision your Azure cloud environment | Platform Tracks → Azure → Prepare |
Provision your AWS cloud environment | Platform Tracks → AWS → Prepare |
Onboard onto SaaS | SaaS → Onboarding and Provisioning |
Understand a topic in depth
If you need to... | Read |
|---|---|
Assess the security and compliance posture | Security & Compliance Brief |
Plan infrastructure costs and capacity | Cost Model Brief |
Understand integration patterns and data flows | Integration Architecture Brief |
Solve a specific challenge
If you need to... | Use |
|---|---|
Implement a non-standard deployment model | Variant Deployment Playbook |
Promote from dev/PoC to production | Environment Promotion Playbook |
Harden the platform before go-live | Day-2 Hardening Playbook |
Handle a production incident | Incident Response Playbook |
The playbook and brief titles above are illustrative — actual content will be linked as it becomes available.