I am a strong advocate for Weaviate Enterprise Cloud, but there are areas where improvement would make a real difference. Monitoring and observability could be more robust out-of-the-box. Currently, I rely on external tools such as Grafana to track my cluster performance, and having a native dashboard with deeper query-level insights would be beneficial. I would appreciate SDK parity across languages. Some newer features are available on the Python SDK before they reach Go and TypeScript, which slows down teams working on other languages. The learning curve for advanced configuration, sharding strategies, replication, and tuning schema design can be steep for newer team members, so better-guided workflows or templates would help. Multi-region support is also a pending request for Weaviate to seamlessly join cross-region platforms. Auto-scaling granularity could be smarter. The current scaling responds to overall resource usage, but it would be better if it could scale independently based on query load versus ingestion load, as these spike at different times for me. Backup and disaster recovery flows need to be more flexible. While backups exist, setting up a cross-cloud failover or point-in-time recovery to a specific transaction can still be manual. Native re-ranking integration has been improved, and these are areas where Weaviate needs continued improvement.
