Hierarchical Supervisor
Managers over leads over specialists, mirroring an org chart.
Problem
A single orchestrator breaks down once a task spans many domains. It needs to understand every specialist’s tools and output, its context fills with their details, and it can’t review dozens of results well. Large jobs need delegation to happen at more than one level.
- Each manager can only coordinate a handful of agents well.
- Domain leads understand their specialists better than a top-level manager does.
- Every extra level adds latency and a chance to lose information.
- Responsibility must stay clear when something goes wrong.
Solution
Arrange agents in a tree. A top supervisor owns the overall goal and talks only to a few domain leads. Each lead owns one area, breaks its part into tasks for its specialists, checks their output and reports a summary upward. Specialists do narrow work with narrow tools. Information flows down as goals and constraints and up as summaries and exceptions. Each level checks the work of the level below before passing it on.
Diagram
Each supervisor talks to 2–5 agents at most. Summaries flow up and exceptions escalate; raw specialist output stays at its own level.
When to use
Use it when
- The task spans several distinct domains, each with its own tools and checks
- One orchestrator would need to manage more than about 5–7 agents
- Different teams own different domains and their agents
- You need clear accountability for each part of the result
Avoid it when
- A flat Orchestrator–Workers setup handles the task fine
- Latency matters; each level adds a round trip
- Domains are so intertwined that leads would constantly need each other
Trade-offs
Gains
- Scales to tasks no single orchestrator could manage
- Each lead's context stays focused on its own domain
- Clear ownership, which matters for audits and debugging
Costs
- Highest cost and latency of the multi-agent patterns
- Information loss through each summary step
- Rigid; cross-domain issues have to travel up and back down
Worked example
An enterprise agent prepares a response to a 60-question security and compliance questionnaire from a prospective customer.
- SupervisorSplits the questionnaire into three areas and assigns each to a lead: Security, Privacy and Legal, and Infrastructure.
- Security leadGives 24 questions to 3 specialists covering encryption, identity and incident response. It reviews their answers against the current policy documents.
- Privacy leadFinds that one question about data residency in a new region conflicts with what the Infrastructure lead reported. It escalates the conflict to the Supervisor instead of guessing.
- SupervisorAsks the Infrastructure lead to confirm. The region launched last month, so the Privacy answer is updated.
- SupervisorAssembles the final response with all 60 answers, flags 4 for human sign-off, and lists which lead owns each section.
Failure modes
| Failure | Symptom | Mitigation |
|---|---|---|
| Telephone game | Key detail lost as summaries move up the tree | Pass structured results with links to source evidence, not just prose |
| Cross-branch conflicts | Two leads give contradictory answers | Leads escalate conflicts; the supervisor runs a consistency check |
| Over-hierarchy | Three levels where one or two would do | Add a level only when a supervisor exceeds its span of control |
| Unclear ownership | Errors bounce between levels with no owner | Every output carries the agent and lead that produced it |