Beta Access: RINNAI is currently in beta on Solana Devnet. View Roadmap →

Solana-native AI Automation Layer

Autonomous Agents.
Deterministic Execution.

RinnaiLabs transforms user intent into secure, non-custodial on-chain execution via autonomous agents. Build automated workflows that execute deterministically with verifiable policy enforcement.

Execution Flow
1
Intent
User defines goal
2
Plan
AI generates execution path
3
Policy
Validate constraints
4
Execute
On-chain enforcement

Contract Address

oPo5Hnw8stVs7ebjg7AkAHsgTbTpcL5kPcdqPTJpump

Explorer

Security Notice: Never trust unofficial contract addresses. Always verify through official RinnaiLabs channels before interacting.

The Automation Problem

Current blockchain automation requires technical expertise, poses security risks, and lacks intelligent orchestration.

Fragmented Automation

Dozens of isolated bots requiring manual configuration and monitoring across different protocols.

80%
of users manage 5+ bots

Manual Workflows

Complex multi-step operations demand constant attention and precise timing to execute correctly.

60%
time spent monitoring

Intent Disconnect

Users forced to learn specific commands and syntax instead of expressing natural goals and constraints.

3.5x
higher error rate

Custody Risk

Traditional bots require handing over private keys, creating single points of failure and security vulnerabilities.

100%
trust dependency

* Illustrative statistics based on industry research

How RINNAI Solves It

An AI automation layer that combines intelligent orchestration with deterministic on-chain enforcement.

1. Understand Intent

Natural language processing interprets user goals, constraints, and desired outcomes without requiring technical syntax.

2. Plan Execution

AI generates an optimal execution path considering gas costs, liquidity, timing, and protocol interactions.

3. Validate Policy

Policy engine checks constraints: spending limits, allowed programs, slippage tolerance, and risk parameters.

4. Submit Request

Agent constructs and submits a signed execution request to the on-chain program with all validated parameters.

5. On-Chain Enforcement

Solana program verifies signatures, validates PDAs, and enforces all policy constraints immutably on-chain.

6. Deterministic Result

Transaction either succeeds within bounds or fails safely with clear reason. No partial states or custody risk.

Design Principles

Core tenets that guide every architectural decision in RINNAI.

Intent over Instructions

Users express what they want to achieve, not how to achieve it. The system handles complexity, optimization, and technical details automatically.

AI without Custody

Agents generate execution plans but never control funds. All transactions require user signatures and are enforced by on-chain programs.

Deterministic Execution

Despite AI's probabilistic nature, execution is guaranteed deterministic through policy validation and program-enforced boundaries.

Composable by Default

Every policy, intent pattern, and agent template can be reused, extended, and composed into more complex automation workflows.

Security-First Architecture

Multi-layer security: policy validation, PDA verification, on-chain enforcement, and complete audit trails for every execution.

Execution Flow

A visual timeline of how intent transforms into deterministic on-chain execution.

User Defines Intent

Natural language goal with constraints

01
"Swap 10 SOL to USDC when price > $150"

AI Generates Plan

Optimal execution path with protocols

02
Jupiter aggregator → Best route → Transaction params

Policy Engine Validates

Check all constraints and boundaries

03
Max spend ✓ | Allowed programs ✓ | Slippage ✓

Agent Submits Request

Signed execution request to program

04
Transaction data + User signature + Policy hash

On-Chain Program Enforces

Immutable validation and execution

05
Verify PDA → Check policy → Execute or revert

Transaction Completes

Deterministic success or safe failure

06
Audit trail logged → User notified → Verifiable

Core Modules

Five interconnected modules that power secure, intelligent automation on Solana.

Intent Engine

Natural language processing and intent parsing. Converts user goals into structured execution parameters with context awareness.

Agent Orchestration

Coordinates multiple AI agents for complex workflows. Handles parallel execution, dependencies, and retry logic automatically.

Policy Engine

Validates all constraints before execution. Supports spending limits, program allowlists, time windows, and custom rules.

Execution Program

On-chain Solana program using PDA-based enforcement. Verifies signatures, validates policies, and executes deterministically.

Audit & Trace Layer

Complete execution history with full verifiability. Every intent, plan, policy check, and result permanently recorded.

System Architecture

User Layer
Intent Input
AI Layer
Planning & Policy
Blockchain Layer
On-Chain Enforcement

Use Cases

RINNAI enables automation across every category of blockchain activity.

Automated Limit Orders

Execute swaps when price conditions are met without constant monitoring.

Yield Optimization

Automatically rebalance positions across protocols for maximum APY.

Dollar-Cost Averaging

Periodic purchases with customizable schedules and market conditions.

Build with RINNAI

Start building deterministic automation in minutes with our SDK.

import { RinnaiClient, defineIntent, attachPolicy } from '@rinnai/sdk';

const client = new RinnaiClient({ cluster: 'devnet' });

// Define user intent
const intent = defineIntent({
  goal: 'Swap 10 SOL to USDC when price > $150',
  constraints: {
    maxSlippage: 0.5,
    timeout: '1h'
  }
});

// Attach policy
const policy = attachPolicy(intent, {
  maxSpend: 10,
  allowedPrograms: ['Jupiter', 'Raydium'],
  requireConfirmation: true
});

// Simulate and execute
const plan = await client.simulatePlan(policy);
const result = await client.submitExecution(plan);

Roadmap

Our journey to production-ready deterministic automation on Solana.

Phase 1

Core Infrastructure

Q1 2025
  • Intent engine with NLP parsing
  • Policy engine with constraint validation
  • Execution program on Solana devnet
  • Basic agent orchestration
  • Developer SDK and documentation
  • Playground for testing
Phase 2

Marketplace & Templates

Q2 2025
  • Agent marketplace launch
  • Pre-built intent templates
  • Protocol integration partners
  • Advanced policy DSL
  • Multi-agent workflows
  • Mainnet deployment
Phase 3

Protocol & Governance

Q3 2025
  • Deep protocol integrations
  • Governance token launch
  • Agent quality scoring
  • Cross-chain bridging support
  • Enterprise features
  • Auditing and compliance tools

Frequently Asked Questions

Common questions about how RINNAI works and ensures security.

Build Deterministic Automation on Solana

Transform user intent into secure, verifiable on-chain execution. Start building with RINNAI today.