A session is a grouping of traces based on a session ID attribute. When building or debugging an LLM chatbot application, being able to see groups of messages or traces belonging to a series of interactions between a human and the AI can be particularly helpful.
This tutorial covers how to add sessions and users as attributes to spans for LLM tracing to assist in finding where conversations break and groups of traces where an application is not performing well (i.e. LLM hallucinations). It also covers how to leverage LLM-assisted evaluations find best/worst performing sessions and users.
Docs: https://docs.arize.com/arize/large-language-models/sessions-and-users
Notebook covering how to implement sessions in real-time: https://colab.research.google.com/github/Arize-ai/tutorials_python/blob/main/Arize_Tutorials/Tracing/Arize_Tutorial_OTLP_Tracing_Llama_Index.ipynb
🔗 Other Links:
Follow Aman Khan on X: https://x.com/_amankhan
More about LLM-assisted evaluation: https://arize.com/blog-course/llm-evaluation-the-definitive-guide/
Join the Arize community: https://join.slack.com/t/arize-ai/shared_invite/zt-26zg4u3lw-OjUNoLvKQ2Yv53EfvxW6Kg
