Hybrid Search for Enterprise AI 

Native Vector Search: Maximize generative AI response accuracy by finding the most relevant data. 

You can now store, index and query vector embeddings to enable generative AI use cases, deep learning and similarity analysis. Use the Optic API to include vectors in multi-model queries to find the best search results. 

BM25 Relevance Ranking: Boost full-text search recall with better document ranking to return the most relevant search results. 

You can now use BM25 as a relevance-scoring algorithm in lexical or hybrid queries. Easily combine with vector queries, merge scores and rerank the results via Optic API for the best quality LLM outputs.   

Graph Shortest Path: Solve real-world problems related to route planning, network optimization, and resource allocation. 

You can now leverage the Shortest Path graph algorithm to quickly traverse a knowledge graph and identify the shortest paths between entities or concepts. This information can be used to provide additional context for LLMs in RAG systems. 

Flexible Analytics and Data Exploration 

Virtual Views: Run large-scale and ad-hoc analytics cost-effectively with relational views generated at query time.

You can now create a relational view of unstructured information on demand to enable analysts to analyze and explore data with familiar SQL-based business intelligence tools in real time. Virtual Views allow you to decrease the storage footprint of large-scale analytical queries and speed up the delivery cycle of ad-hoc reports during data discovery.   

Optimized Scalability and High Availability

Dynamic Hosts: Handle changing workloads to deliver consistent user experiences during peak demand.  

When you anticipate periods of high query demand, we’ve made it easy to scale up your clusters during usage spikes and scale them down when loads are back to normal. This helps you deliver consistent performance while optimizing resource utilization. 

Hardened Security and Encryption

Enhance your security posture by taking advantage of the latest encryption algorithms and network security protocols.

MarkLogic Server 12 ships with support for OpensSSL 3.3, TLS 1.3 and and federal Information Processing standards (FIPS) 140-3. The latest TLS protocol introduces security enhancements designed to fortify network communications against modern threats. OpenSSL 3.3 introduces cutting-edge cryptographic algorithms, enhanced security defaults and compliance with the latest FIPS requirements.

Modernized Platform Support

Modernize your technology stack and deploy MarkLogic Server in the environments that best fit your strategy.

MarkLogic server 12 comes with support for Red Hat Enterprise Linux 9, Amazon Linux 2023, Windows 2022 and "rootless" Docker Images. It also includes an upgrade of the embedded Javascript engine to Google V8 12.4, and updated document filters for content and metaData extraction from 1000+ binary formats.

MarkLogic Server 12 is available under enterprise and developer license. For more information about the full list of new feature and enhancements, please see the release notes.