Skip to content Skip to sidebar Skip to footer

How Agent-Ready Search Schemas Can Make RAG More Accurate and Cheaper

What Happened Vespa 8.786.55 can expose schema guidance to AI agents through /skills/v1/ when the schema skill is enabled in its service configuration. Teams can add explanations to schemas, fields, and rank profiles; Vespa incorporates those explanations into guidance that changes with the schema [1]. In a Vespa evaluation of 196 data-dependent tasks and variants,…

Read More

How to Improve RAG Search Quality Without Paying to Rerank 200 Results

What Happened A comparison across 25 selected retrieval benchmarks found that reranking more results did not reliably improve search. With Elasticsearch BM25 as the initial retriever, reranking 50 candidates reduced error by 18.75%; reranking 100 reduced it by 20.97% but used twice the reranking tokens. The largest datasets were excluded from this comparison [1]. Initial…

Read More

How to Build RAG Search That Enforces Document Access Without Reindexing

What Happened Two Elasticsearch capabilities address different parts of the retrieval problem. ES|QL Data Federation can query Parquet, CSV, TSV and NDJSON files in Amazon S3 without copying or indexing them, and can query those datasets alongside Elasticsearch indices. It requires an Enterprise subscription and is experimental in Elastic Cloud Hosted 9.5 and in technical…

Read More

How to Build RAG Search That Finds the Right Evidence Without Exposing the Wrong Data

What Happened Kibana now offers more than 90 AI connectors, including Slack, GitHub, Confluence, PagerDuty, Gmail, and Google Drive. An agent can call these services as tools alongside Elasticsearch queries, rather than requiring a custom integration for each service or copying all external data into a search index [1]. That makes it possible to connect…

Read More

How to Cut Vector Search Storage Without Sacrificing Retrieval Quality

What Happened DiskBBQ changed how it auto-calibrates compressed vector indexes. Its sweep now puts more weight on vector size, tests preconditioning for every candidate, and can favor deeper reranking when the combination meets the same recall target. The cost model changed from document_bits + 1.3 × rerank_depth to 3.0 × document_bits + 1.3 × rerank_depth…

Read More

How to Make RAG Search Faster, Cheaper and More Trustworthy

What Happened Qdrant 1.20 adds two quantization-native vector datatypes, turbo8 and turbo16, plus advanced indexing for nested metadata. These target the memory and disk costs of vector storage and the speed of filtered search. Qdrant reports performance and cost improvements, but production results will depend on each workload’s vectors, filters and recall requirements [1]. Elastic…

Read More

Weaviate’s GraphQL Retirement: How to Keep RAG Search Working Through the Migration

What Happened Weaviate deprecated its GraphQL API on October 26, 2026. Existing deployments do not break immediately, but new Weaviate Cloud clusters already have GraphQL disabled. Cloud clusters with no GraphQL usage in the preceding 30 days will have it disabled on December 1, 2026; all remaining Cloud clusters follow on May 1, 2027. Requests…

Read More

Weaviate Is Retiring GraphQL: How to Keep RAG Search Working During Migration

What Happened Weaviate deprecated its GraphQL query API on October 26, 2026, without immediately disrupting existing users. New Weaviate Cloud clusters already have GraphQL disabled. Cloud will disable it on December 1, 2026 for clusters with no GraphQL use in the preceding 30 days, and on May 1, 2027 for all remaining clusters [1]. Applications…

Read More

How to Build More Reliable RAG Search with Hybrid Retrieval, OCR and Cited Answers

What Happened Recent releases and implementations show RAG search moving beyond vector similarity alone. Qdrant 1.20 adds BM25 ranking over full-text payloads without requiring embeddings, a tokenizer for code search, faster indexing for nested filters, and vector-storage options including a 16-bit format that halves storage without reported accuracy loss [1]. An Elasticsearch demonstration turned a…

Read More

Can 32K-Token Embeddings Replace Chunking in RAG? How to Choose the Right Retrieval Design

What Happened DMIS Lab’s UltraBERT-Embed-150M is a 151M-parameter English embedding model that accepts inputs up to 32K tokens and produces one 768-dimensional vector per input. That makes a single-vector representation possible for many documents that would otherwise be split into chunks. A Qdrant test compares document-level vectors, 512-token chunks, a combination of both, and compressed…

Read More

Engineering Founded & Customer Focused

KIMBODO LLC. © 2026. All Rights Reserved.