Build AI Apps With Guardrails
A framework for building with security, design consistency, prompt management, and real data. Works with Cursor, Windsurf, Antigravity, and all your favorite tools.
64% Less Code
Same functionality, dramatically simpler code
import { ChatOpenAI } from "@langchain/openai";
import { HumanMessage, SystemMessage } from "@langchain/core/messages";
const model = new ChatOpenAI({
modelName: "gpt-4",
temperature: 0.7,
});
const messages = [
new SystemMessage("You are helpful."),
new HumanMessage("Hello!"),
];
const response = await model.invoke(messages);
console.log(response.content);import { createRana } from '@rana/core';
const rana = createRana();
const response = await rana.chat('Hello!');
console.log(response.content);Zero to Production in 5 Minutes
From first command to production deployment in under 5 minutes
Install
One command to get started. No complex setup.
npx create-rana-app my-ai-appBuild
Write clean, simple AI code with full TypeScript support.
const rana = createRana();
const response = await rana.chat('Hello!');Test
Built-in testing framework for AI applications.
aiTest('works', async () => {
await expect(response).toSemanticMatch('greeting');
});Deploy
One-click deploy to Vercel, AWS, or any platform.
vercel deployThree Powerful Packages
Production-ready AI building blocks that work together or standalone
@rana/helpers
10 one-line AI functions for common tasks
const summary = await summarize(longText);
const spanish = await translate(text, { to: 'es' });
const category = await classify(text, ['spam', 'ham']);@rana/prompts
Enterprise prompt management with A/B testing
const pm = new PromptManager({ workspace: 'app' });
await pm.register('greeting', { template: '...' });
const result = await pm.execute('greeting', vars);@rana/rag
Advanced RAG with hybrid retrieval & re-ranking
const pipeline = RAGPresets.balanced();
await pipeline.index(documents);
const result = await pipeline.query({ query: '...' });The Guardrails You Need
A framework that works WITH your tools to enforce security, consistency, and quality at every step
Security Guardrails
OWASP Top 10 protection, GDPR compliance, PII detection, and prompt injection defense built into every request.
Design System Enforcement
Automatically enforce your design library usage and component patterns. No more inconsistent UIs.
Prompt Management
Version control for prompts, A/B testing, analytics, and team collaboration. Manage AI behavior like code.
Real Data Only
Framework prevents fake/mock data in production. Connect to real databases and APIs with type-safe queries.
Controlled Features
Quality gates prevent unauthorized feature additions. Every change must meet your standards before deployment.
Works With Your Tools
Use Rana with Cursor, Windsurf, Google Antigravity, Claude Code, or any IDE. Complementary, not competitive.
Zero Vendor Lock-in
Switch between 9 LLM providers in one line. OpenAI, Anthropic, Google, xAI, Mistral, Cohere, Together, Groq, Ollama.
70% Cost Savings
Automatic cost reduction through intelligent model routing, caching, prompt optimization, and budget enforcement.
Loved by Developers
Join thousands of developers building production AI apps with RANA
“RANA's cost tracking alone saved us $50k/month. The built-in testing gave us confidence to ship faster.”
“Coming from LangChain, RANA felt like a breath of fresh air. We shipped in a quarter of the time.”
“The PII detection and audit logging made HIPAA compliance straightforward. We launched months ahead of schedule.”
“RANA's automatic fallbacks and circuit breakers mean our students never experience downtime.”
“The intelligent model routing automatically picks the cheapest model that meets our quality bar. Magic.”
“Finally, a framework that treats testing as a first-class citizen. Our CI pipeline actually works now.”
“We evaluated every AI framework. RANA was the only one that had everything we needed out of the box.”
“The MCP server creation feature alone was worth switching. Standardized our entire AI integration layer.”
“90% less code is not an exaggeration. Our RAG pipeline went from 200 lines to 20.”
Ready to build your AI agent?
Start building production-ready AI agents in minutes with our comprehensive training and documentation. Need expert help? Waymaker offers implementation services.