What a blast at @n8n_io In The Loop! The past two days were super dense of discussions around @qdrant_engine, vector search and real-world use cases ranging from e-commerce search to internal knowledge bases, from fast and accurate retrieval at scale to agentic memory. A few takeaways for me: - People care about privacy and data residence, and Qdrant is the perfect solution: not only it's OSS and self-hostable at any scale, but we also offer Private Cloud, deployable on-prem, air-gapped, suitable for the most sensitive enterprise use cases - Search workloads are reaching a critical mass: many people ask how they can cut storage and costs, and Qdrant helps here as well, thanks to our effort to bring quantization-native storage data type (such as turbo4), in order to cut storage without sacrificing too much recall - Knowledge bases for personal/company usage are becoming more and more common, especially given the increasing share of work that agents handle for us. Qdrant here offers the perfect layer to store and retrieve your documents with high accuracy and low latency, while being extremely easy to integrate via SDK, API or even n8n node :) I also gave a talk about a self-healing agent learning to avoid its previous mistakes by storing and retrieving human feedback in Qdrant, and you can find the slides here (the recording will be available in the next few weeks!): qdrant-talks.clelia.dev/Buil… Thanks to n8n for this amazing event and, if you were around and I didn't get a chance to talk to you, please feel free to connect with me and send a DM!✨

One of the things I did for @n8n_io In The Loop was to self-host n8n on my Kubernetes cluster to showcase my @qdrant_engine demo. I learnt a good chunk of hands-on k8s from this experience, especially in debugging/understanding failure modes that single-node applications never exposed me to, but also on the data (Postgres) and networking (@Cloudflare) side. I wrote about it in my latest blog post: clelia.dev/blog/2026-10-25-s… Hope you like it!✨