Building blocks for AI agents

Your business logic, from day one.

Lua is like LEGO for AI agents. Snap together pre-built components — LLMs, integrations, channels, data — and focus entirely on what makes your business unique. No infrastructure. No boilerplate. Just building.

Channels

  • WhatsApp
  • Slack
  • Web chat

AI models

  • Claude
  • GPT-4
  • Gemini

Integrations

  • Salesforce
  • Stripe
  • PostgreSQL

Your logic · day 1

  • The part only you can write

The result

  • Production AI agent

The difference

Build, don't configure.

Spend your time on business logic, not plumbing.

Without LuaWith Lua
Set up Kubernetes clusters for auto-scalingAuto-scaling handled automatically
Manage multiple LLM provider APIs and rate limitsOne interface for all LLM providers
Build custom data pipelines with Airbyte + DagsterBuilt-in RAG with semantic search
Configure vector databases for embeddingsVector embeddings managed for you
Write integrations for each messaging channelDeploy to all channels from one codebase
Handle token optimization and cost trackingCost optimization built-in
Months of setup before first deploymentShip in hours, not months
traditional-setup.sh15+ services
# Traditional AI agent setup

# 1. Set up Kubernetes cluster
kubectl apply -f kubernetes/
kubectl apply -f ingress/
kubectl apply -f monitoring/

# 2. Deploy data infrastructure
helm install airbyte airbyte/airbyte
helm install neo4j neo4j/neo4j
helm install redis bitnami/redis

# 3. Set up orchestration
docker-compose up -d dagster cube

# 4. Install AI dependencies
pip install langchain openai anthropic
pip install chromadb pinecone-client

# 5. Channel integrations
npm install @slack/bolt
npm install whatsapp-web.js
npm install instagram-private-api

# 6. Configure everything...
# ... 3-6 months later ...
lua-setup.sh3 commands
# Lua setup — that's it

npm install -g lua-cli

lua init my-agent

lua deploy


# Auto-scaling
# Multi-channel (web, Slack, WhatsApp...)
# LLM routing (OpenAI, Anthropic, Gemini)
# RAG pipeline
# Data ingestion
# Monitoring & logs

# You're live. Ship in hours.

Under the hood

The platform powering your agent.

A sophisticated infrastructure you never have to think about.

LuaYour single interface to everything below
AgentsSkills & tools400+ integrationsMulti-channelJobs & webhooksMemoryRAG

Infrastructure we manage

Auto-scaling

Elastic compute on demand

LLM routing

Multi-provider switching

Data ingestion

Universal connectors

Data lake

Version-controlled storage

Knowledge graph

Semantic relationships

Orchestration

Workflow automation

Semantic layer

Unified metrics

API gateway

Global edge delivery

Channels

Multi-platform messaging

Integrations

400+ services ready

Cache layer

Multi-tier performance

Security

Enterprise-grade protection

Monitoring

Real-time observability

What you actually write

Just your business logic in TypeScript.

tools/checkInventory.ts
import { z } from "zod";
import { LuaTool } from "lua-cli";

export const checkInventory: LuaTool = {
  name: "check_inventory",
  description: "Check if a product is in stock",

  input: z.object({
    productId: z.string().describe("The product ID"),
  }),

  async execute({ productId }) {
    const response = await fetch(
      `https://api.mystore.com/inventory/${productId}`
    );

    const data = await response.json();

    return {
      inStock: data.quantity > 0,
      quantity: data.quantity,
      warehouse: data.location
    };
  }
};

15+

Services abstracted

400+

Enterprise integrations

600+

Data connectors

6

Channels supported

Data intelligence

Your data, automatically understood.

Connect any data source. Our AI figures out what it means.

Traditional data pipelines require you to define schemas, write transformations, and maintain complex ETL jobs. Lua's data platform loads first and understands later — AI interprets your data, infers relationships, and builds a knowledge graph automatically.

01

Your data

Salesforce, HubSpot, Stripe, PostgreSQL...

02

Ingest

600+ connectors, zero config.

03

AI interprets

Schema inference, entity resolution.

04

Knowledge graph

Relationships + vectors.

05

Your agent

Instant access, natural language.

Continuous improvement

Watch your AI get smarter.

Tell us what matters. We'll measure everything, automatically.

You say

“Track customer happiness and whether we actually solve their problems”

Live dashboard

Happiness score8.4/10
Problems solved73%
Response quality92%
↑ Trending up this week

Plain English · AI-powered · Auto charts · Real-time · Smart suggestions · Trends & insights

Enterprise ready

Connect your entire stack.

Your AI agents work seamlessly with the tools your business already uses.

CRM & sales

Salesforce · HubSpot · Pipedrive · Zoho

Communication

WhatsApp · Slack · Discord · Telegram · Gmail · Outlook

Databases & storage

PostgreSQL · MongoDB · AWS S3 · Google Cloud · Redis

Payments & finance

Stripe · PayPal · QuickBooks · Xero

Productivity & support

Zendesk · Jira · Notion · Asana · Trello · Shopify

And 400+ more integrations available out of the box.

Why Lua

Everything you need to build.

Pre-built blocks.

Every component you need is ready to use. Channels, LLMs, integrations, data pipelines — just snap them together.

Day 1 productivity.

Write business logic from minute one. No setup, no configuration, no waiting. Your first agent ships today.

Plug & play.

400+ integrations ready to connect. Swap LLMs, add channels, connect databases — all without rewriting code.

Your logic only.

Focus on what makes your business unique. The checkout flow, the approval process, the customer journey — that's all you write.

Scale automatically.

From prototype to production without changing a line. Built on enterprise infrastructure that scales with you.

Iterate fast.

Test locally, deploy instantly, rollback safely. Experiment with new features without risk.

Start building today.

Your first AI agent is three commands away. What will you build?

npm install -g lua-cli

lua init my-agent

lua deploy