Stack AI

Stack AI handles RAG natively. When you upload documents or connect a data source, the platform automatically indexes your content and creates a Knowledge Base node. Connect it to an LLM node and the system runs vector-based semantic search behind the scenes, delivering context-rich, cited responses with as few as 4 nodes.

Stack AI supports OpenAI, Anthropic, Google Gemini, Meta Llama, xAI, Mistral, Perplexity, Groq, Cerebras, TogetherAI, Replicate, Azure OpenAI, AWS Bedrock, and local LLMs via custom endpoints. All models are available without requiring separate API keys or external accounts.

Yes. Stack AI offers on-premise deployment and Virtual Private Cloud (VPC) options for organizations with strict data residency or sovereignty requirements. This is available on the Enterprise plan and can be deployed on AWS, Azure, or GCP using Kubernetes.

Stack AI provides native, OAuth-based connections with 100+ tools including Microsoft SharePoint, Salesforce, Workday, HubSpot, Jira, Slack, Oracle NetSuite, and Amazon S3. No API keys or technical configuration is needed—users authenticate through standard SaaS login flows.

Stack AI provides granular RBAC, admin approval workflows for publishing agents, version control, automatic audit logs of all inputs and outputs, token usage tracking, SSO integration with Okta and Entra ID, and built-in guardrails for prompt and response filtering. The platform is SOC 2 Type II, HIPAA, GDPR, and CCPA compliant with ISO 27001 certification.