Build AI agents in minutes
Learn to create AI agents with RAG, knowledge bases, LLM integrations, and built-in guardrails for reliable and secure applications.
What's in this video
- Create an AI agent and connect it to a knowledge base.
- Add RAG and an LLM integration to the workflow.
- Apply built-in guardrails before using the application.
Frequently asked questions
Topics from this video and why governed AI matters.
- What types of agents can you build in SUPERWISE?
- Two main types: an AI assistant retrieval agent for chat (single data source, context-specific responses) and an advanced agent for complex processes (multiple tools). You add context (e.g. Wikipedia or SQL), connect an LLM, add a prompt, then save and publish with guardrails.
- How do guardrails work when building agents?
- Guardrails are rule-based policies that ensure safe, accurate, and appropriate content (docs.superwise.ai): rules, ethical guidelines, and content filters. They prevent harmful or misleading outputs and can restrict topics (e.g. block “How old is Taylor?”). You combine one or more rules, test in a playground, and use them natively when deploying or via API/SDK, so the check runs before the answer is returned.
- Why add guardrails before deploying an agent?
- Agents that answer from a knowledge base can drift off-topic or leak sensitive content. Guardrails block or audit those cases before they reach users. Governed AI requires that you enforce safety at build time and at runtime.
- Why is governed AI essential for RAG and knowledge-base agents?
- RAG agents mix user input and retrieved context. Without guardrails and observability, you cannot control what gets retrieved or returned. SUPERWISE gives you rules, visibility, and audit so you trust what the agent says and does.
Related Videos
Sentinel: AI systems your auditors already trust
Meet the SUPERWISE Sentinel AI gateway, which redacts sensitive data, applies guardrails, and reports what passed, changed, or was blocked.
demo
SUPERWISE® Council builds governed AI systems
See SUPERWISE Council turn plain-English requirements into an AI architecture, visual system map, versioned guardrails, tests, and deployment gates.
demo
How Claude builds SUPERWISE agents for you (MCP demo)
See Claude Code use SUPERWISE API and SDK documentation over MCP to build an agent, attach guardrails, and create working notebooks.
demo
SUPERWISE v1.27: self-service Starter tier and full guardrails
Review the SUPERWISE Starter tier announcement, including built-in guardrails and Remote MCP support for deploying and governing AI agents.
update
How SUPERWISE stops patient data leaks
Watch SUPERWISE detect a patient's Social Security number and block the PII in real time, before it ever reaches the LLM.
vertical
SUPERWISE v1.25.0: AI agents, guardrails, and observability
Review SUPERWISE v1.25.0 features for multi-agent systems, including security controls and observability for monitoring AI agents.
update