Easy RAG logo
Back to Blog
Company

Introducing EasyRAG

We're launching EasyRAG to make RAG accessible to every developer. Here's our story.

Vlad Racoare
4 min read
AnnouncementProduct

Introducing EasyRAG: RAG as a Service

Today, we're excited to launch EasyRAG - the simplest way to add RAG (Retrieval-Augmented Generation) to your applications. Upload documents, search semantically, and get AI-powered answers with just a few lines of code.

The Problem We're Solving

Building RAG systems is hard. You need to:

  1. Set up vector databases (Pinecone, Weaviate, Qdrant)
  2. Handle document processing (PDFs, DOCX, etc.)
  3. Chunk text optimally
  4. Generate embeddings (OpenAI, Cohere)
  5. Implement search and retrieval
  6. Connect to LLMs for generation
  7. Handle errors, rate limits, and costs
  8. Scale infrastructure

That's weeks of work before you even start building your actual product.

Our Solution

EasyRAG handles all of this for you. Here's everything you need:

javascript
import { EasyRAG } from '@easyrag/sdk';

const client = new EasyRAG(process.env.EASYRAG_API_KEY);

// 1. Upload documents
await client.upload('my-dataset', [
  file1.pdf,
  file2.docx,
  file3.csv
]);

// 2. Search semantically
const results = await client.search(
  'my-dataset',
  'What is the refund policy?'
);

// 3. Get AI answers
const answer = await client.query(
  'my-dataset',
  'Summarize our Q3 performance'
);

console.log(answer.data.result);
// "Based on your documents, Q3 revenue was $2.1M..."

That's it. Production-ready RAG in 10 minutes.

What Makes EasyRAG Different?

1. Dead Simple API

Three methods to learn: upload(), search(), query(). That's it.

No complex configuration, no infrastructure to manage, no ML expertise needed.

2. Multiple File Formats

We handle the complexity of different file types:

  • Documents: PDF, DOCX, XLSX, PPTX, TXT, MD, CSV
  • Media: MP3, WAV, MP4 (with auto-transcription)

Upload anything, we'll process it.

3. Smart Chunking

We automatically chunk your documents for optimal retrieval. Need custom chunking? You can configure it:

javascript
await client.upload('dataset', file, {
  chunkSize: 500,      // Tokens per chunk
  chunkOverlap: 50     // Context overlap
});

4. Flexible Metadata

Tag your documents with any metadata and filter on it:

javascript
// Upload with metadata
await client.upload('dataset', file, {
  metadata: {
    'document.pdf': {
      department: 'legal',
      year: 2024,
      confidential: false
    }
  }
});

// Filter searches
const results = await client.search('dataset', query, {
  filters: [
    { key: 'department', match: { value: 'legal' } },
    { key: 'year', match: { value: 2024 } }
  ]
});

Perfect for multi-tenant apps, user isolation, or document classification.

5. Streaming Responses

Real-time responses for better UX:

javascript
for await (const chunk of client.queryStream('dataset', question)) {
  if (chunk.delta) {
    process.stdout.write(chunk.delta);
  }
}

Users see answers as they're generated - no waiting.

6. Built for Production

  • Multi-tenant: Isolate customer data with filters
  • Scalable: We handle infrastructure
  • Fast: <100ms search latency
  • Reliable: 99.9% uptime SLA
  • Secure: SOC 2 compliant (in progress)

Use Cases

Customer Support Bots

javascript
// Auto-answer support tickets from your docs
const answer = await client.query(
  'support-kb',
  'How do I reset my password?'
);

await sendResponse(ticket, answer.data.result);

Internal Knowledge Base

javascript
// Help employees find information instantly
const results = await client.search(
  'company-docs',
  'What is our PTO policy?'
);

Research Assistants

javascript
// Search through research papers
const results = await client.search(
  'research-papers',
  'Recent advances in transformer architectures'
);

Document Analysis

javascript
// Extract insights from contracts
const answer = await client.query(
  'legal-docs',
  'Find all termination clauses in these contracts'
);

Content Recommendations

javascript
// Find similar articles
const related = await client.search(
  'blog-posts',
  article.content
);

Pricing

We believe in simple, transparent pricing:

Free Tier

  • 100 free credits
  • Upload: 1 credit per file
  • Query: 0.1 credit per question
  • Perfect for prototyping

Pay As You Go

  • $10 for 1,000 credits
  • $45 for 5,000 credits
  • $200 for 25,000 credits
  • No monthly fees, no surprises

Enterprise

  • Volume discounts
  • Dedicated support
  • Custom SLAs
  • Contact us

Getting Started

1. Sign Up

Create your free account at easyrag.com.

2. Get Your API Key

Generate an API key from the dashboard.

3. Install the SDK

bash
npm install @easyrag/sdk

4. Start Building

javascript
import { EasyRAG } from '@easyrag/sdk';

const client = new EasyRAG(process.env.EASYRAG_API_KEY);

// Your RAG app is ready!

Check out our Quick Start Guide for detailed instructions.

Roadmap

We're just getting started. Coming soon:

Q1 2025:

  • React components library
  • Real-time collaboration features
  • Advanced analytics dashboard
  • Webhook support

Q2 2025:

  • Fine-tuning support
  • Custom embedding models
  • Multi-language support
  • GraphQL API

Q3 2025:

  • On-premise deployment
  • Advanced security features
  • Batch processing API
  • ML model marketplace

Want something specific? Let us know!

Why We Built This

As developers, we've built RAG systems multiple times. Each time, we spent weeks setting up infrastructure, debugging edge cases, and optimizing performance.

We realized 90% of RAG implementations need the same thing: upload documents, search them, get AI answers. The complexity is unnecessary.

EasyRAG is the RAG system we wish existed when we started.

Our Mission

Make RAG accessible to every developer.

Whether you're building a solo project or a startup serving millions, you deserve simple, powerful tools. No ML PhD required.

Join Us

We're a small team backed by [investors who believe in our vision]. We're:

  • Obsessed with developer experience
  • Committed to transparent pricing
  • Building in public
  • Always listening to feedback

Stay Updated

Resources

Try It Now

Ready to add RAG to your app? Start building in 5 minutes:

  1. Sign up for free
  2. Follow the Quick Start
  3. Build something amazing

Have questions? We're here to help:

What's Next?

Over the coming weeks, we'll be publishing:

  • Technical deep-dives on RAG architecture
  • Case studies from early customers
  • Best practices for production RAG
  • Tutorials for common use cases

Subscribe to our newsletter to stay updated.

Acknowledgments

Thank you to:

  • Our early beta users for invaluable feedback
  • The open-source community (LangChain, Qdrant, OpenAI)
  • Our investors who believed in the vision
  • Everyone who helped us get here

Let's make RAG easy together. 🚀

Ready to start? Sign up now and get 100 free credits.