AI Avatar knowledge base
for builders

Connect your avatar to product docs, sales collateral, course material or support guides, so it can search your content during a conversation.

Built for teams and developers

Brand Guidelines

“Our company uses a clear, friendly voice and consistent branding across every channel.”

Product docs

“Create your first avatar, connect your knowledge sources, and start a conversation.”

Sales collateral

“Turn product questions into conversations. Guide every prospect with answers tailored to their needs.”

Pricing Info

“Compare plans, included features, usage limits, and billing options to find the right fit.”

Talk with Caleb

Ground your avatar in your own content.

Anam’s knowledge tools retrieve relevant passages from your documents for the avatar to use in its answers. Add clear instructions and test responses against your sources.

1

Ingest

Ingest

Create a knowledge folder and upload PDF, TXT, Markdown, DOCX, CSV or JSON files.

2

Attach

Attach

Connect a knowledge tool to your avatar and choose its folders.

3

Run

Run

Start a real-time session using your avatar ID. The avatar can search the connected documents to answer relevant questions.

Unmatched realism.
Try a grounded avatar.

Talk to Caleb, an avatar connected to Anam’s product docs. Ask about the platform and try a knowledge-backed conversation with Cara-4.1 Express.

Caleb, an Anam AI avatar

Brand Guidelines

Product docs

Sales collateral

Pricing Info

const base = "https://api.anam.ai/v1";
async function post(path, body) {
  const r = await fetch(base + path, {
    method: "POST",
    headers: {
      Authorization: `Bearer ${process.env.ANAM_API_KEY}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify(body),
  });
  if (!r.ok) throw new Error(`Anam request failed: ${r.status}`);
  return r.json();
}
const tool = await post("/tools", {
  name: "search_product_docs",
  description: "Search our product docs for features and setup.",
  type: "SERVER_RAG",
  config: { documentFolderIds: [process.env.ANAM_KNOWLEDGE_GROUP_ID] },
});
const persona = await post("/personas", {
  name: "My first Anam avatar",
  avatarId: process.env.ANAM_AVATAR_ID,
  voiceId: process.env.ANAM_VOICE_ID,
  llmId: "a7cf662c-2ace-4de1-a21e-ef0fbf144bb7",
  systemPrompt: "Answer product questions with search_product_docs. " +
    "Use only relevant results. Say when you cannot find an answer. " +
    "Keep replies to one or two short sentences.",
  toolIds: [tool.id],
});
const { sessionToken } = await post("/auth/session-token", {
  personaConfig: { personaId: persona.id },
});
const base = "https://api.anam.ai/v1";
async function post(path, body) {
  const r = await fetch(base + path, {
    method: "POST",
    headers: {
      Authorization: `Bearer ${process.env.ANAM_API_KEY}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify(body),
  });
  if (!r.ok) throw new Error(`Anam request failed: ${r.status}`);
  return r.json();
}
const tool = await post("/tools", {
  name: "search_product_docs",
  description: "Search our product docs for features and setup.",
  type: "SERVER_RAG",
  config: { documentFolderIds: [process.env.ANAM_KNOWLEDGE_GROUP_ID] },
});
const persona = await post("/personas", {
  name: "My first Anam avatar",
  avatarId: process.env.ANAM_AVATAR_ID,
  voiceId: process.env.ANAM_VOICE_ID,
  llmId: "a7cf662c-2ace-4de1-a21e-ef0fbf144bb7",
  systemPrompt: "Answer product questions with search_product_docs. " +
    "Use only relevant results. Say when you cannot find an answer. " +
    "Keep replies to one or two short sentences.",
  toolIds: [tool.id],
});
const { sessionToken } = await post("/auth/session-token", {
  personaConfig: { personaId: persona.id },
});

Connect your models and knowledge

Preply learners practising with interactive avatars
Preply learners practising with interactive avatars
Preply

Preply saw 44% higher student retention with Anam

Preply uses Anam’s real-time avatars to give learners more opportunities to practice speaking and build confidence between lessons.

Built for your most important conversations.

SOC 2 Type II

Move from pilot to production with independently audited security controls, backed by Anam’s SOC 2 Type II report.

HIPAA

Bring a human face to patient education, intake and clinical training, with security and privacy controls that support HIPAA-compliant deployments.

GDPR

Build for European audiences with privacy safeguards, clear data-processing commitments and support for GDPR requirements.

Frequently asked questions

How do I connect a knowledge base to my avatar?

In Anam Lab, open your avatar’s Prompt tab and upload documents through the Knowledge card. Anam creates and attaches a knowledge tool on the first upload. Describe when it should search, wait for the files to be ready, then publish your avatar and test a new conversation.

Which files can I add to the knowledge base?

You can upload PDF, TXT, Markdown, DOCX, CSV and JSON files. Use clear, readable text and keep related facts and exceptions together. Documents become searchable once processing is complete; PDFs need extractable text, not just scanned images.

Can I update the knowledge base after launch?

Yes. Upload the revised file, wait until it is ready, then remove the superseded version and test the changed information in a fresh conversation. Document updates do not require republishing the avatar. Attaching a new knowledge tool in Lab does.

Can I use my own custom avatar with a knowledge base?

Yes. Create a custom avatar from a compatible photo or AI-generated image in Lab or through the API, then select it for your avatar. The same avatar can use your chosen voice, prompt and knowledge tools. Changing its appearance does not require rebuilding the knowledge base.

Can I use a custom LLM or my existing AI agent?

Yes. Register a compatible streaming LLM in Anam Lab or through the API and select it for your avatar. To use Anam’s knowledge tools, the model must also support tool calling. If your application runs the agent itself, it controls retrieval and sends the resulting responses to Anam.

Can my avatar respond while it looks up information?

Yes. With multi-utterance turns, your avatar can say “Let me check that for you,” search its knowledge base, then continue with the answer in the same turn. This keeps the conversation feeling natural while information is retrieved. It requires an LLM that supports interleaved tool calling. Anam-provided models hosted by OpenAI or Azure support this; Groq-hosted models do not.

What if the answer is not in my documents?

Use the avatar’s prompt to tell it to acknowledge missing information, ask a clarifying question or direct the user to a person who can help. Knowledge retrieval does not guarantee a correct answer, so test questions your documents can answer as well as ones they cannot.

How long can an avatar conversation last?

The maximum session length depends on your plan and the avatar’s configured time limit. You can set a shorter limit for your use case. Check your organization’s current limits in Anam Lab and end sessions when they are no longer needed.

Can several people talk to my avatar at once?

Yes, within your organization’s simultaneous-session allowance. Each conversation uses one session slot, and websites using the same organization share that allowance. Plan capacity around the busiest times and how long conversations last.