Siddhant Kulkarni is a researcher affiliated with New York University who specializes in multi-agent systems and orchestration architectures for large language models. Kulkarni has contributed to the field through empirical research on agent coordination patterns, particularly focusing on the technical trade-offs between different orchestration approaches.
Kulkarni's notable contribution to the field includes a comprehensive benchmark evaluating multi-agent orchestration patterns. This research involved conducting empirical testing across a substantial dataset of 10,000 documents, examining performance characteristics across five frontier and open-weight large language models (LLMs). The study was designed to systematically understand the architectural trade-offs inherent in different approaches to orchestrating multiple AI agents 1).
The benchmark assessed four distinct multi-agent orchestration patterns, providing empirical evidence about their relative strengths, weaknesses, and performance characteristics across different model scales and architectures. By testing on both frontier proprietary models and open-weight alternatives, the research offers insights applicable to practitioners with varying resource constraints and deployment preferences.
Kulkarni's research addresses a critical area in multi-agent AI systems: determining which orchestration architectures are most suitable for different use cases and constraints. Multi-agent orchestration patterns represent different methodologies for coordinating multiple AI agents to solve complex problems—including sequential patterns, hierarchical patterns, and other coordination strategies that affect system reliability, efficiency, and capability.
The empirical nature of this work—grounding conclusions in testing on thousands of documents rather than purely theoretical analysis—reflects a practical orientation toward understanding how these systems perform in realistic scenarios. The inclusion of both frontier models and open-weight alternatives suggests the research addresses concerns relevant to both resource-intensive enterprise deployments and more cost-constrained implementations.
Kulkarni's institutional affiliation with New York University places the researcher within one of the major centers for AI and machine learning research in the United States. This affiliation provides access to research resources, collaboration opportunities, and the academic infrastructure that supported the large-scale empirical evaluation of orchestration patterns.