Quick Start

This guide will help you get started with Denser Retriever in just a few minutes. You'll learn how to install the SDK, create a knowledge base, ingest documents, and perform semantic searches.

Installation

npm install @denserai/retriever-sdk

pip install denser-retriever-sdk

Initialize the Client

First, initialize the Denser Retriever client with your API key:

import { DenserRetriever } from "@denserai/retriever-sdk";

const client = new DenserRetriever({ apiKey: "YOUR_API_KEY" });

from denser_retriever import DenserRetriever

client = DenserRetriever(api_key="YOUR_API_KEY")

Never hardcode your API key in production code. Use environment variables instead.

Complete Workflow

Follow these steps to build your first semantic search application:

Create a Knowledge Base

Knowledge bases are containers for your documents. Create one to organize related content:

const kbRes = await client.createKnowledgeBase(
  "Product Documentation", // Name
  "All product-related docs" // Description
);
const kbId = kbRes.data.id;
console.log(`Created knowledge base: ${kbId}`);

kb = client.create_knowledge_base(
    name="Product Documentation",
    description="All product-related docs"
)
kb_id = kb["data"]["id"]
print(f"Created knowledge base: {kb_id}")

Ingest Documents

You can ingest documents in two ways: file upload or raw text import.

Option A: Import Raw Text

The simplest way to add content:

await client.importTextContentAndPoll(
  kbId,
  "Getting Started Guide", // Title
  "Denser Retriever is a semantic search platform that enables..." // Content
);

client.import_text_content_and_poll(
    knowledge_base_id=kb_id,
    title="Getting Started Guide",
    content="Denser Retriever is a semantic search platform that enables..."
)

Option B: Upload Files (PDF, DOCX, etc.)

For file uploads, use the 3-step process:

import fs from "fs";
import axios from "axios";

// 1. Get presigned upload URL
const fileName = "manual.pdf";
const stats = fs.statSync("./manual.pdf");
const presignRes = await client.presignUploadUrl(kbId, fileName, stats.size);
const { fileId, uploadUrl } = presignRes.data;

// 2. Upload file to S3
const fileStream = fs.createReadStream("./manual.pdf");
await axios.put(uploadUrl, fileStream, {
  headers: {
    "Content-Type": "application/octet-stream",
    "Content-Length": stats.size,
  },
});

// 3. Import and wait for processing
const result = await client.importFileAndPoll(fileId);
console.log("Document status:", result.data.status);

import os
import requests

file_path = "./manual.pdf"
file_size = os.path.getsize(file_path)
file_name = os.path.basename(file_path)

# 1. Get presigned upload URL
presign = client.presign_upload_url(
    knowledge_base_id=kb_id,
    file_name=file_name,
    size=file_size
)
file_id = presign["data"]["fileId"]
upload_url = presign["data"]["uploadUrl"]

# 2. Upload file
with open(file_path, "rb") as f:
    requests.put(upload_url, data=f, headers={"Content-Type": "application/octet-stream"})

# 3. Import and wait for processing
result = client.import_file_and_poll(file_id)
print("Document status:", result["data"]["status"])

Search Your Documents

Perform semantic search across your knowledge base:

const searchRes = await client.query("How do I get started?", {
  knowledgeBaseIds: [kbId], // Optional: filter by knowledge bases
  limit: 5, // Optional: limit results (default: 10, max: 50)
});

searchRes.data.forEach((item) => {
  console.log(`[Score: ${item.score.toFixed(3)}] ${item.content}`);
});

results = client.query(
    query="How do I get started?",
    knowledge_base_ids=[kb_id],  # Optional: filter by knowledge bases
    limit=5,  # Optional: limit results (default: 10, max: 50)
)

for item in results["data"]:
    print(f"[Score: {item['score']:.3f}] {item['content']}")

Knowledge Base Management

List Knowledge Bases

const listRes = await client.listKnowledgeBases();
listRes.data.forEach((kb) => {
  console.log(`${kb.name} (${kb.id})`);
});

kbs = client.list_knowledge_bases()
for kb in kbs["data"]:
    print(f"{kb['name']} ({kb['id']})")

Update Knowledge Base

await client.updateKnowledgeBase(kbId, {
  name: "Updated Documentation",
  description: "New description",
});

client.update_knowledge_base(
    knowledge_base_id=kb_id,
    name="Updated Documentation",
    description="New description"
)

Delete Knowledge Base

await client.deleteKnowledgeBase(kbId);

client.delete_knowledge_base(kb_id)

Deleting a knowledge base will permanently remove all associated documents. This action cannot be undone.

Account Management

Check Usage

const usage = await client.getUsage();
console.log(`Knowledge bases: ${usage.data.knowledgeBaseCount}`);
console.log(`Storage used: ${usage.data.storageUsed} bytes`);

usage = client.get_usage()
print(f"Knowledge bases: {usage['data']['knowledgeBaseCount']}")
print(f"Storage used: {usage['data']['storageUsed']} bytes")

Check Credit Balance

const balance = await client.getBalance();
console.log(`Credit balance: ${balance.data.balance}`);

balance = client.get_balance()
print(f"Credit balance: {balance['data']['balance']}")

Next Steps

Check out the TypeScript SDK for detailed TypeScript documentation
Check out the Python SDK for detailed Python documentation
Browse the API Reference for all available endpoints

Quick Start

On this page