Platform Capabilities

Agentic data workflows powered by leading LLMs. Build intelligent applications with pre-configured AI models and seamless integration capabilities.

• Multi-model AI orchestration and intelligent routing
• Pre-built agentic workflows for common tasks
• Context-aware AI responses with memory management
• Real-time model performance optimization

Raw sensitive data stays in your environment. Only redacted or transformed tokens may be sent to approved external models, depending on your configuration. Enterprise-grade security with comprehensive compliance controls.

• Raw sensitive data stays in your environment. Only redacted or transformed tokens may be sent to approved external models, depending on your configuration
• End-to-end encryption for all data in transit and at rest
• Regular security audits and penetration testing
• Designed to support GDPR, DORA, NIS2 and EU AI Act requirements when deployed and configured accordingly. ISO 27001 and SOC 2 certifications are in progress.

Maintain complete control over your data with flexible deployment options and compliance with regional data protection regulations.

• Regional data residency with geo-specific deployment
• Data localization for GDPR, CCPA, and other regulations
• On-premise and private cloud deployment options
• Data isolation and multi-tenancy with strict boundaries

Collections are curated, verified knowledge bases that ground AI responses in trusted information. Unlike ChatGPT or Claude, which attach files as conversation context limited by context windows, Aimable processes your documents through a RAG pipeline — indexing them in a vector database for semantic search and a graph database for understanding relationships between documents. The result: the AI retrieves the most relevant passages across thousands of documents, with source citations pointing to the exact files used.

• Personal Collections for individual users and Shared Collections managed by designated Curators with RBAC controls
• Curators can instruct AI to update, correct, or expand collection content — making collections living, continuously improving knowledge bases, not static file dumps
• Indexed via vector database (meaning-based retrieval) and graph database (document relationships) for accurate retrieval at scale
• Data Sources: named connections to external storage (Google Drive folders, SharePoint document libraries, or direct upload) that feed files into the RAG pipeline automatically. GitHub and Dropbox planned
• Citation-backed responses with links to source documents so you can verify before you act

Detect and handle sensitive data before it reaches external LLMs. Three configurable levels per Space: Automatic (PII redacted in real-time), Human-in-the-loop (PII flagged, user decides), or Off (for local models with zero data egress).

• Three levels per Space: Automatic redaction, Human-in-the-loop review, or Off for local models
• Real-time detection of PII, financial data, and sensitive information
• Configurable redaction rules and patterns per organisation
• Comprehensive audit logs of all redaction activities

Ensure AI outputs meet quality, safety, and compliance standards. Configurable guardrails prevent harmful content, and the upcoming Policy Checker adds a compliance layer that checks conversations against your corporate policies.

• Content filtering for harmful or inappropriate output
• Bias detection and mitigation in AI responses
• Custom policy enforcement for brand safety
• Real-time quality assurance and validation
• Policy Checker (in development): a dedicated Policy Q&A Space where users ask questions grounded in actual policy documents, plus cross-space Policy Monitoring that flags potential violations without blocking conversations

Every interaction — chat or API — is tracked. The audit trail captures prompts, responses, documents retrieved (with RAG relevance scores), compliance events, latency, token usage, and user metadata. Configurable per Space with two logging levels.

• Full logging (every interaction) or events-only logging (blocks, redactions, policy violations) — configured per Space
• Documents retrieved with RAG hits and relevance scores, so you can see what informed each answer
• Compliance events: PII redactions, policy flags, and blocked queries — all recorded with context
• Latency and token usage metrics per interaction, user, and team for cost tracking
• Logs exportable for GDPR Article 30 compliance reporting and SIEM integration

Secure ChatGPT alternative with enterprise governance built-in. Chats take place within a Space and inherit all its configuration. They are private to the user unless placed in a shared Project.

• Familiar ChatGPT-like interface with enterprise security
• Admin-controlled System Prompt per Space shapes AI behaviour and tone — users cannot modify it, ensuring consistent compliance
• Configurable logging level per Space: full logging (every interaction) or events-only (blocks, redactions, policy violations)
• Integration with Collections and role-based access controls

Five defined roles govern who can do what across the platform. Permissions are enforced consistently across chat, API, and collection management.

• Owner — full platform administration across all Spaces and settings
• Admin — manage Spaces, users, and Space-level configuration
• Curator — manage shared Collection content: add, edit, remove files, and instruct AI to update knowledge bases
• Auditor — view audit trails and compliance reports without access to modify data or configuration
• User — use assigned Spaces, create Projects and Chats within their permissions

OpenAI-compatible API and MCP (Model Context Protocol) integrations for connecting your automations to governed AI. API keys are scoped to a Space — all Space configuration (models, data access, PII rules) applies to programmatic access. Organisations can create dedicated API Spaces with tailored settings for automation.

• OpenAI-compatible API — drop-in replacement for existing integrations
• MCP (Model Context Protocol) integrations for connecting AI tools and agents
• API keys scoped to a Space — same models, PII rules, and data access apply as for chat users
• Works with n8n, Make, Zapier, and custom applications
• Both chat and API interactions flow through the same audit trail
• Cost tracking per user and team built-in