R2 Advisory

Products

iDragonFly

An emerging AI-native operating environment for governed expert work.

In Development

Who It Is For First

iDragonFly is not built for every knowledge worker on day one. The initial focus is expert teams managing complex, multi-step work where context and authority must persist.

  • Consulting and advisory teams managing multi-step client work
  • Research teams coordinating evidence, agents, and analysis
  • Product teams running agent-assisted workflows
  • Cyber and investment teams where context and authority must persist across a long-running engagement

The Problem

Current AI tools place the burden of coordination and context management on the user.

The user must manually manage context, files, prompts, models, agent handoffs, approval, memory, status, evidence, and decision history. Every new chat starts over. Every agent operates with a different slice of context. Expert work needs persistent organizational memory, and it needs agents, models, data, tools, humans, and authority coordinated against that memory, not reassembled by hand every time.

What iDragonFly Coordinates

Eight things that currently do not talk to each other.

People

AI agents

Models

Tools

Knowledge

Authority

Workflows

Organizational memory

The burden today: Context, Files, Prompts, Models, Agent handoffs, Approval, Memory, Status, Evidence, Decision history.

Product Thesis

An active coordination layer for governed expert work.

  • Project Manager Agents that maintain objectives and status
  • Persistent knowledge that survives across sessions and agents
  • Model routing across the right tool for each task
  • Governed delegation, built on ADC principles
  • Human approval at defined checkpoints
  • Traceable execution and organizational memory
  • Integration with enterprise systems

Illustrative Product Workflow, Not Yet Implemented

A market-entry plan, coordinated end to end.

A consulting or internal strategy team is developing a market-entry plan. A Project Manager Agent maintains objectives and status. Research agents gather and evaluate evidence. Analysis agents test options. Human experts approve assumptions and decisions. The system preserves sources, rationale, tasks, and authority. Deliverables and implementation actions remain connected to the working knowledge base.

Relationship to ADC

ADC is open. iDragonFly is commercial. The Agent Delegation Contract informs how authority and delegation work inside iDragonFly: what an agent operating inside the system is actually permitted to decide and do, and what it must be able to prove. iDragonFly includes coordination functions beyond ADC, such as project memory, model routing, and human approval workflows.

Product Status

What design partners can help validate.

iDragonFly is in development. This page does not represent a claim of general availability, broad adoption, production maturity, or established customers.

Initial workflowsGovernance requirementsIntegration prioritiesKnowledge and memory modelUser experienceEnterprise deployment constraints

Related Thesis

The Enterprise Memory Problem

The underlying argument for why AI-native organizations need persistent memory infrastructure: the problem iDragonFly is designed to address.

Read on MichaelERuiz.com →

Explore a design partnership.

If your organization is wrestling with agent coordination, enterprise memory, or governed expert work, we would welcome the conversation.

Explore a Design Partnership