AITOPAI热报
How-toSelected

Milvus Function Chain Reranks Search with XGBoost in One Call

Milvus@milvusio
Post
View the post
Why we picked it

Reranking inside the vector database itself removes the need for a separate reranking service and an extra network hop, which simplifies production search stacks. Builders running e-commerce, feed, or RAG systems can reuse the open-source demo to blend relevance with revenue-driving signals. It compares with the common pattern of exporting candidates from Milvus and reranking them in an external model server.

Milvus demonstrated its Function Chain feature, which combines semantic similarity with business signals such as price, freshness, popularity, and ratings in product search. The pipeline then applies an XGBoost model to rerank the results, all executed inside Milvus within a single search call. The team published a live demo at demos.milvus.io and open-sourced the code on GitHub. The same approach can extend to news feeds, community content, and knowledge base search.

Image · Milvus
Original · Milvus

In product search, the closest semantic match isn’t always the best result once price, freshness, popularity, and ratings enter the picture. In this demo, we show how 𝗠𝗶𝗹𝘃𝘂𝘀 𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻 𝗖𝗵𝗮𝗶𝗻 𝗰𝗼𝗺𝗯𝗶𝗻𝗲𝘀 𝘀𝗲𝗺𝗮𝗻𝘁𝗶𝗰 𝘀𝗶𝗺𝗶𝗹𝗮𝗿𝗶𝘁𝘆 with business signals such as price, freshness, popularity, and rating. It then uses an XGBoost model to rerank the results—all inside Milvus and in a single search call. You can also apply the same approach to news feeds, community content, knowledge bases, and other search systems where relevance i demos.milvus.io/function-chain… 🔗 Try the live demo: https://t github.com/zc277584121/mi… ore the code on GitHub: https://t.co/yDdf0tIO5f Your browser does not support the video tag. 🔗 View on Twitter 💬 0 🔄 0 ❤️ 0 👀 60 ⚡

#Milvus#XGBoost#RAG#vector database

Keep reading

Recent stories with the same tag or section
How-toBuildersSelected23:50
Zilliz tests Jev model as semantic cache gateway guard

Semantic caching usually forces a tradeoff between cost savings and serving wrong answers to slightly different queries; a cheap guard model shifts that balance toward accuracy. Engineers running RAG or agent workloads can plug this pattern into existing caches like GPTCache to cut LLM spend without degrading quality. Milvus and Zilliz have active developer communities across ASEAN, so the benchmark and blog are directly accessible to teams in the region.

#Milvus#Zilliz#Jev#semantic caching
Milvus@milvusioSource
How-toBuildersSelected00:11
Milvus benchmark tests Jev against LLMs in agentic search

The benchmark suggests a practical routing rule: swap general-purpose LLM calls for Jev only on narrow, high-frequency decisions like stopping checks, and keep dedicated rerankers like Voyage rerank-3 for re-ranking. Teams running RAG pipelines can use the latency and cost figures to decide where Jev cuts spend without hurting recall. Full results are on Milvus's blog.

#Jev#Milvus#RAG#Voyage
Milvus@milvusioSource
How-toBuildersSelected00:30
Milvus 3.0 demo shows late-interaction retrieval for visual PDFs

The practical shift is that dense retrieval no longer has to flatten a page into one vector, so queries about a single label or arrow can still hit the right region. Teams indexing engineering manuals, papers, financial reports or slide decks can try the live Milvus demo to see whether late interaction beats their current single-vector pipeline on visually dense documents.

#Milvus#late interaction#RAG#vector search
Milvus@milvusioSource