Actian Analytics Engine 8.0: Data Lake Enabled and AI-Ready

Actian Analytics Engine 8.0 delivers a complete, high-performance foundation for the modern data stack, bridging the gap between mission-critical analytics and the emerging requirements of AI. This release activates your data for the modern ecosystem by integrating native support for open table formats, streamlined developer workflows, and the foundational building blocks for Generative AI. Rather than acting as a static repository, version 8.0 serves as a dynamic compute layer that connects to, caches, and processes data across your entire environment. 

Actian Analytics Engine 8.0 is built for sovereignty across infrastructure, data, and AI and allows for an air-gapped modern architecture within your own premises.

A Note on Our Evolution: 

As we continue to modernize our stack, Actian Vector has officially evolved into the Actian Analytics Engine. This change aligns with the rebranding of our managed service, now the Actian Analytics AI Platform. While these offerings are tailored for different deployment models—on-premises, private cloud, and fully managed—they share the same core technology, technical range, and ground-to-cloud capability.  

Bringing AI to Your Data

A central focus of version 8.0 is enabling AI capabilities directly within the user database environment, reducing the need to move large datasets to external AI platforms. While model training typically occurs in external environments, version 8.0 allows you to bring those models into the engine to run and manage them where your data resides.  

• Native VECTOR Support: Version 8.0 adds native support for the VECTOR data type and functions for similarity search. 
• Unified Vector Storage: Unify relational and VECTOR data in a single system, allowing for strong consistency and efficient query processing without maintaining multiple specialized databases.  
• In-Database RAG & LLM Access: The engine serves as a high-performance knowledge base for Retrieval-Augmented Generation (RAG). Utilizing native GenAI UDFs, you can prompt LLMs directly over SQL, augmented with your local relational data. 
• In-Database ML Inference: In addition to Tensorflow, you can now execute model inference as part of your analytics by leveraging support for ONNX models. 
• Feature Preprocessing & Metrics: Built-in functions for common stages of the ML workflow, including scaling and encoding, as well as metrics to monitor model performance. 
• Standardized AI Connectivity: A dedicated MCP (Model Context Protocol) Server ensures that modern AI agents and development tools can interact with your data through a standardized interface. 

Activating the Data Lakehouse

Analytics Engine 8.0 breaks down the walls between the warehouse and the data lake, enabling you to query live data sets across your entire distributed landscape.  

• Open Table Format Integration: Simplified integration of external catalogs, including support for Iceberg, Hudi, and Delta Lake. 
• Transparent Caching: To ensure superior performance on repetitive data lake queries, the engine automatically caches external data sources to create the experience of working with internal tables. 
• Simplified Registration: The new CREATE DATA SOURCE command abstracts external warehouses/DBMSs or CloudFS buckets into a single DDL query with automated type mapping. 

Streamlining the Developer & ETL Experience

This release reduces friction for those building applications or managing complex data pipelines.  

• Transient Table Support: Designed to accelerate ETL and temporary use cases across session boundaries, transient tables reside in RAM for higher throughput and lower latency, and are automatically dropped when the server stops. 
• Vector Python SDK: A new Python SDK available via standard package managers simplifies building applications using the Python Database API Specification v2.0. 
• Data Science Workbench: Spin up browser-based environments conveniently with  
• Jupyter notebook support, VS Code integration, and pre-configured connectivity to the engine. 
• Open Data Observability: All-new Actian Monitor 1.0, built on an OpenTelemetry (OTel) based monitoring system, provides deep telemetry via Grafana and Prometheus. 

Performance & Foundation

Innovation remains anchored in the raw performance that defines our engine.  

• Breakthrough Performance: 8.0 delivers significant gains, with TPC-DS 1TB benchmarks showing execution times up to 2.5x faster than previous versions. 
• Runtime Memory Scaling: Adjust memory configuration (Bufferpool vs. Query Memory) during runtime to adapt to shifting workloads without system downtime. 
• Hardened Security: Communication is secured by default with TLS 1.3, and new  
• Credential Management: DDL commands manage secrets securely by name.

Learn more about Actian Analytics Engine 8.0 at Actian Analytics Engine (Formerly Vector):

About Author

About Douglas Dailey

Douglas Dailey is Director of Product Management for Actian's cloud and on-prem databases, tools, and connectivity solutions. Over 15 years, Doug has built data virtualization platforms and steered technology investments in IBM Netezza, DB2 Replication, and Informix. He's led sessions at major data events (e.g., IBM's Think conference) and authored Whitepapers on data hub and data fabric topologies. On the Actian blog, Doug focuses on hybrid data strategies, replication, and emerging use cases. Check out his latest articles to understand modern data architectures.