Parallelization

Parallelization is effective when the divided subtasks can be parallelized for speed, or when multiple perspectives or attempts are needed for higher confidence results.

The concurrent orchestration pattern runs multiple AI agents simultaneously on the same task.

This workflow involves breaking down a task into independent subtasks that can be handled concurrently by multiple LLM calls or agents. Outputs are then programmatically aggregated and synthesized into a result.

Claude Research

For speed, we introduced two kinds of parallelization: (1) the lead agent spins up 3-5 subagents in parallel rather than serially; (2) the subagents use 3+ tools in parallel. These changes cut research time by up to 90% for complex queries, allowing Research to do more work in minutes instead of hours while covering more information than other systems.

There's no clear conflict resolution strategy to handle contradictory or conflicting results from each agent.

Running multiple agents in parallel can increase immediate resource utilization and token consumption, which leads to higher operational costs. Furthermore, the gather step requires complex logic to synthesize potentially conflicting results, which adds to the development and maintenance overhead of the system.