flashserve/RAGPulse
An Open-Source RAG Workload Trace to Optimize RAG Serving Systems
This project provides a realistic dataset of user interactions with a Q&A system that uses Retrieval Augmented Generation (RAG). It includes anonymized records of user questions, system prompts, retrieved information, and chat history. AI infrastructure engineers can use this data to test and improve the performance of their RAG-based systems.
Use this if you need to optimize the speed and efficiency of your RAG serving infrastructure using real-world usage patterns.
Not ideal if you are looking for raw conversational text or a dataset for training RAG models, as all content is replaced with anonymized hash IDs.
Stars
35
Forks
2
Language
Python
License
MIT
Category
Last pushed
Nov 18, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/flashserve/RAGPulse"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
OpenBMB/UltraRAG
A Low-Code MCP Framework for Building Complex and Innovative RAG Pipelines
Quansight/ragna
RAG orchestration framework ⛵️
microsoft/rag-time
RAG Time: A 5-week Learning Journey to Mastering RAG
AnkitNayak-eth/EpsteinFiles-RAG
A RAG pipeline implementation built on the 'Epstein Files 20K' dataset from Hugging Face (Teyler).
apify/apify-haystack
The official integration for Apify and Haystack 2.0