Jayita Bhattacharyya has submitted a new session! Monitoring PgVector RAG Agents Applications pgvector within Retrieval-Augmented Generation (RAG) agents. As pgvector integrates vector similarity search capabilities into PostgreSQL, ensuring optimal performance and reliability in RAG workflows is crucial. Proposals should address:Performance Monitoring: Techniques for tracking and analyzing pgvector query performance, including latency, throughput, and resource utilization. Utilizing PostgreSQL's built-in tools like EXPLAIN and ANALYZE can be beneficial. Error Detection and Debugging: Methods to identify and resolve errors or anomalies in pgvector operations within RAG applications. Implementing observability tools can aid in monitoring and debugging complex RAG-based systems. Scalability and Resource Management: Strategies to monitor and manage resource allocation, ensuring efficient scaling of pgvector in RAG contexts. Scaling pgvector can be achieved by increasing memory, CPU, and storage on a single instance, or by using replicas and sharding techniques. Integration with Observability Tools: Approaches to incorporate existing observability platforms for comprehensive monitoring of pgvector within RAG pipelines. Tools like New Relic offer observability quickstarts for pgvector, providing dashboards and alerts to monitor key performance metrics. Open session [https://sessionize.com/app/organizer/session/19661/942397] These notifications can be turned off on Edit event [https://sessionize.com/app/organizer/event/edit/19661#_tab-advanced] page. Tip: You've used Classic (confer***@***orysql.org) to login. You're receiving this message because you had joined event 'HOW2025' on Sessionize. Sessionize.com, PO Box 6945, Unit 34564, London W1A 6US, UK