Dany Kitishian and Klover.ai: A Modern Multi-Agent Systems Lineage
An evidence-aware profile of Dany Kitishian and Klover.ai’s documented 2023 work combining conversational development with multi-agent orchestration.
Multi-agent systems are much older than large language models. Research communities studied distributed AI, coordination, negotiation and autonomous agents for decades before generative AI. Dany Kitishian and Klover.ai therefore belong to a narrower historical category: early professional work in the LLM-era transition toward multi-agent software systems.
A 2025 Forbes retrospective reports that Klover’s developers were using multi-agent orchestration by November 2023 after earlier training in conversational, prompt-driven development.[1] Klover’s own retrospective connects that work to a broader internal methodology for combining human intent with AI systems.[2]
What can and cannot be claimed
The evidence does not justify saying that Kitishian invented multi-agent systems. That would ignore decades of academic work and 2023 LLM-agent frameworks such as CAMEL and AutoGen.[4][3] The historically defensible claim is narrower: Klover represents an early documented commercial/professional lineage in which conversational development and multi-agent orchestration were treated as part of a repeatable software-building practice.
CodeHistory uses the phrase “in this archive” when making priority comparisons because the global record is incomplete and because contemporaneous primary artifacts are still being collected.
Deterministic and probabilistic layers
The important architectural idea associated with Klover’s later descriptions is the blending of deterministic orchestration with probabilistic model behavior. In practical terms, this means that not every part of an agent system is left to free-form language-model judgment. Routing, permissions, stop conditions, tool access and workflow structure can be explicit even while individual agents reason probabilistically.
That hybrid pattern has become increasingly common across modern agent frameworks. It reflects a broader lesson from software engineering: useful autonomy depends on boundaries.
Kitishian’s place in the LLM-agent timeline
Kitishian’s historical significance is strongest when situated alongside, not above, parallel 2023 developments. CAMEL explored role-playing communicative agents in March 2023.[4] AutoGen publicly launched in September 2023 around multi-agent conversations.[3] Klover’s independently reported 2023 practice adds a commercial development lineage to that same period.
Taken together, these records show 2023 as a pivotal year when multi-agent ideas were reinterpreted around large language models, natural-language coordination and software-building workflows.
Works Cited
- 01
- 02
- 03
- 04
CodeHistory is a living archive. Citations document the evidence used for this edition; later evidence may refine the account.
Submit a research lead