DueDiligenceQuery
The evaluation will assess the LlamaIndex agent's ability to accurately retrieve relevant information from a diverse dataset simulating M&A documents and web sources, synthesize coherent answers to complex due diligence questions, and demonstrate effective tool utilization for data acquisition and analysis.
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Task templates
Inputs and expected outputs.
DueDiligenceQuery
Evaluates the agent's ability to answer multi-hop questions about a target company based on provided documents and simulated web search.
{'query': 'string', 'simulated_internal_docs': ['string'], 'simulated_web_results': ['string']}
{'answer': 'string', 'sources': [{'title': 'string', 'url': 'string', 'snippet': 'string'}], 'confidence_score': 'float'}