I want to build an investigation AI agent, which LLM and framework should I use. Give me…
- User
I want to build an investigation AI agent, which LLM and framework should I use. Give me a simple direct answer and explanation
Assistant
Direct Answer:
- LLM: GPT-4 or Gemini Pro 1.5
- Framework: Langchain
Explanation:
- LLM Choice: GPT-4 and Gemini Pro 1.5 have strong reasoning, knowledge retrieval, and code understanding capabilities, crucial for investigation. Gemini Pro 1.5's large context window is particularly useful for handling extensive documents and data often found in investigations.
- Framework Choice: Langchain provides the modularity and tools (chains, agents, document loaders, etc.) needed to orchestrate a complex investigation workflow. It facilitates connecting the LLM to external data sources, running multiple queries, and building a reasoning chain.
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.