From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company
- Published
- Source
- arXiv
- Paper number
- 164
- Field
- Agents / Multi-Agent
- arXiv ID
- 2604.22446
Key points
- Current multi-agent systems rely on fixed team structures and tightly coupled coordination logic, which makes them inflexible and brittle for novel, open-ended projects.
- Existing frameworks face incompatible runtimes and session-bound learning problems, which hinder continuous knowledge accumulation and interoperability across different agents.
- The lack of a unified organizational abstraction makes it difficult to systematically manage, coordinate, and evolve an AI workforce for complex real-world tasks.
- OneManCompany (OMC) adopts a Talent-Container architecture that integrates heterogeneous agents as employees, with portable identities and an abstraction over diverse runtime environments.
- Project execution is formalized as Explore-Execute-Review (E2R) tree search, providing structured planning, execution, and evaluation, along with formal guarantees of termination and no deadlock.
- OMC implements systematic self-evolution mechanisms at both the individual agent and organization level, including self-reflection, updated standard operating procedures (SOPs), and human resources performance management for continual learning.
Paper links
External research summaries. These are not HDATF publications or measured product results.