Enterprise Agents

5 roles - each with owner, sponsor, and configurable orchestrator/LLM

Enterprise Roles

🚗

Project Driver

Drives multiple projects & products. Delegates, approves.

orchestrator: langgraph llm: gpt-4o
🎯

Product Lead

Owns product direction & roadmap. Makes product decisions.

orchestrator: langgraph llm: gpt-4o
👥

Group Admin

Drives collaboration across groups. Facilitates, engages.

orchestrator: langgraph llm: gemini-2.0-flash
👔

Team Lead

Leads one team. Manages members, coaches, 1:1s.

orchestrator: langgraph llm: gpt-4
👤

Employee Assistant

Supports employees. IT admin defaults, employee overrides.

orchestrator: langgraph llm: gemini-2.0-flash

Create Enterprise Agent

from orchestrator import create_agent

# With owner & sponsor
agent = create_agent("proj-driver", "project_driver", 
    config={
        "owner": "cto@company.com",
        "sponsor": "cfo@company.com",
        "projects": ["proj-a"]
    })

Agent Configuration

Governance
ConfigTypeDefaultDescription
ownerstr-Accountable for outcomes
sponsorstr-Budget/strategy approval
reports_tostr-Reporting chain
priorityint1Priority level
Scope
projectslist[]Projects
productslist[]Products
groupslist[]Groups
manageslist[]Direct reports
Runtime
orchestratorstrlanggraphRuntime
llmstrrole defaultLLM model

Runtime

# Set default runtime (not in code)
export RUNTIME=langgraph