Skip to content

PROACTIVA-US/ROLLIZR

Repository files navigation

Rollizr - AI-Native M&A Platform

Fragmented Industry Rollup Platform (FIRP) powered by Claude AI agents.

Overview

Rollizr is an AI-native platform that orchestrates intelligent agents to streamline M&A deal sourcing, evaluation, and execution across fragmented service industries. Built on Claude, the platform reduces weeks of manual research to minutes through automated workflows for target identification, valuation, compliant outreach, due diligence, and post-acquisition integration.

Core Capabilities

  • 🔍 Target Discovery - Scout agent finds and scores companies against investment thesis
  • 🔗 Entity Resolution - Resolver agent deduplicates and merges company records
  • 📊 Company Profiling - Profiler agent enriches data with operational intelligence
  • 💰 Valuation - Valuation agent estimates value using multiple methodologies
  • ⚖️ Compliance - Compliance agent enforces regulatory requirements
  • 📧 Outreach - Outreach agent generates personalized, compliant communications
  • 📋 Due Diligence - Diligence agent manages document requests and IC memos
  • 🔄 Integration - Integrator agent creates 100-day plans and tracks KPIs

Quick Start

Prerequisites

Installation

# Clone the repository
git clone <repository-url>
cd Rollizr

# Install dependencies
npm install

# Configure environment
cp .env.example .env
# Edit .env and add your ANTHROPIC_API_KEY

Environment Configuration

Create a .env file with:

ANTHROPIC_API_KEY=your_api_key_here
CLAUDE_MODEL=claude-sonnet-4-5-20250929
MAX_TOKENS=4096
TEMPERATURE=0.7

Running Examples

# Run sourcing workflow demo
node examples/demo-sourcing.js

# Run outreach workflow demo
node examples/demo-outreach.js

Architecture

Agent System

Rollizr uses a multi-agent architecture where specialized AI agents handle different aspects of the M&A workflow:

┌─────────────────────────────────────────────────────────┐
│                  Agent Orchestrator                     │
│  (Coordinates workflows and agent communication)        │
└─────────────────────────────────────────────────────────┘
                         │
        ┌────────────────┼────────────────┐
        │                │                │
   ┌────▼────┐      ┌───▼────┐      ┌───▼────┐
   │  Scout  │      │Profiler│      │Valuation│
   │  Agent  │      │ Agent  │      │ Agent   │
   └─────────┘      └────────┘      └─────────┘
        │                │                │
        └────────────────┼────────────────┘
                         │
                    ┌────▼────┐
                    │Compliance│
                    │  Agent   │
                    └─────────┘

Project Structure

Rollizr/
├── src/
│   ├── agents/              # Agent implementations
│   │   ├── base-agent.js    # Base agent class
│   │   └── index.js         # Agent factory
│   ├── config/
│   │   └── agents.js        # Agent configurations & prompts
│   ├── utils/
│   │   └── claude-client.js # Anthropic SDK wrapper
│   ├── workflows/
│   │   └── orchestrator.js  # Multi-agent orchestration
│   └── index.js             # Main entry point
├── examples/                # Example workflows
│   ├── demo-sourcing.js
│   └── demo-outreach.js
├── tests/                   # Tests (coming soon)
├── .env.example             # Environment template
├── package.json
├── CLAUDE.md                # AI assistant guidance
└── README.md

Usage

Single Agent Execution

const { AgentOrchestrator } = require('./src');

const orchestrator = new AgentOrchestrator();

// Execute a single agent
const result = await orchestrator.executeAgent('scout', {
  thesis: { /* investment criteria */ },
  companyData: { /* company info */ }
});

console.log(result.output);

Sourcing Workflow

Complete end-to-end target evaluation:

const result = await orchestrator.executeSourcingWorkflow(
  thesis,      // Investment criteria
  companyData  // Company to evaluate
);

if (result.qualified) {
  console.log(`Score: ${result.summary.score}/100`);
  console.log(`Est. Value: $${result.summary.estimatedValue.midpoint}`);
  console.log('Top Signals:', result.summary.topSignals);
}

Outreach Workflow

Generate compliant, personalized outreach:

const result = await orchestrator.executeOutreachWorkflow(
  companyData,  // Company and contact info
  context       // Previous analysis results
);

if (result.success) {
  console.log('Subject:', result.outreachDraft.output.subject);
  console.log('Message:', result.outreachDraft.output.message_body);
}

Custom Workflows

Build your own agent pipelines:

// Sequential execution
const result = await orchestrator.executePipeline(
  ['scout', 'profiler', 'valuation'],
  initialInput
);

// Parallel execution
const result = await orchestrator.executeParallel(
  ['profiler', 'compliance'],
  companyData
);

Agent Configurations

Each agent is configured with:

  • System Prompt - Defines role, responsibilities, and output format
  • Temperature - Controls creativity (0.1-0.7 depending on agent)
  • Max Tokens - Output length limit
  • Role - Functional category for workflow organization

See src/config/agents.js for detailed agent prompts and parameters.

Workflows

Sourcing Workflow

  1. Scout - Score company against thesis
  2. Profiler - Build detailed profile
  3. Valuation - Estimate value (parallel with Compliance)
  4. Compliance - Check licensure and regulatory status

Outreach Workflow

  1. Compliance - Validate outreach permissions
  2. Outreach - Generate personalized message
  3. (Integration with email platform - coming soon)

Diligence Workflow

  1. Diligence - Generate request list
  2. Diligence - Analyze uploaded documents
  3. Diligence - Produce IC memo

Integration Workflow

  1. Integrator - Create 100-day plan
  2. Integrator - Set up KPI tracking
  3. Integrator - Monitor progress

Development

Adding New Agents

  1. Add agent config to src/config/agents.js
  2. Agent will be automatically instantiated in src/agents/index.js
  3. Create custom workflow in src/workflows/orchestrator.js

Running Tests

npm test  # Coming soon

Development Mode

Set NODE_ENV=development in .env to see detailed agent execution logs.

Roadmap

MVP (Current)

  • ✅ Core agent system
  • ✅ Basic orchestration
  • ✅ Sourcing workflow
  • ✅ Outreach workflow
  • 🚧 Data ingestion connectors
  • 🚧 Web interface

v1 (Next)

  • Entity resolution at scale
  • Database integration (PostgreSQL)
  • CRM sync (HubSpot)
  • Email sending (SendGrid)
  • HVAC vertical pack

v2 (Future)

  • Additional verticals (plumbing, landscaping, dental, MSP)
  • Advanced valuation models
  • Document Q&A with RAG
  • Full integration playbooks

Documentation

Tech Stack

  • AI/LLM: Anthropic Claude (via SDK)
  • Runtime: Node.js
  • Future: PostgreSQL, Redis, Elasticsearch, Next.js

Contributing

This is an early-stage project. Contributions welcome!

License

ISC

Support

For questions or issues, see documentation or create an issue in the repository.


Built with ❤️ and Claude AI

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors