Product Launch: Sovereign Infrastructure, Agentic Stewardship & More
Key Takeaways
- Enterprise data teams are balancing governance rigor and regulatory control with growing pressure for faster answers.
- Organizations need more control over where data and AI run without slowing down the business.
- Actian shipped updates across its portfolio to help close that gap.
- The result is governance that teams can scale and infrastructure they can control.
- AI can deliver faster answers grounded in trusted, governed enterprise data.
Enterprise data teams are balancing demands for:
- Governance that earns trust.
- Sovereignty over where data and AI run.
- Answers the business can act on right away.
When the same problems come up across regulated industries, embedded software teams, and data stewards, we build for them. Here’s what’s new.
Data Management: Trust
AI Analyst now runs in more enterprise environments. Customers can self-host on-premises or in a private cloud, and they can bring their own LLM (BYOLLM) for inference they already operate. BYOLLM works in both SaaS and self-hosted deployments.
Impact: Teams can ask questions and get trusted answers from governed data, even when that data or their preferred LLM must stay in their environment.
Data Steward Agent adds Plan Mode visibility and full result-set coverage. Before a bulk change runs, the agent shows its scope, affected items, fields, estimated duration, and risks, with reasoning for each item. Its reads cover the complete result set within a set call and time budget.
Impact: Stewards review the complete scope of what a change touches before they approve it. Catalog-wide answers cover every asset.
Data Contract Builder has more to work with now. The release extends ODCS v3.1.0 coverage with expanded Fundamentals, Roles, and SLA attributes. Teams can bulk upload the contracts they already have, and the builder supports the same languages as the Actian Data Intelligence Platform. Contract owners also get a daily email digest of pending suggestions across all their contracts.
Impact: Teams move existing contracts in without rebuilding them. Owners see open suggestions in one daily digest.
Lifecycle Policies ensure data assets meet governance requirements, such as descriptions and ownership, before data is available to data consumers, including business users relying on AI. These requirements are enforced across Studio, imports, bulk edits, and APIs.
Impact: Consumers find documented, owned data. Governance teams build metadata quality into the publishing workflow.
Dedicated VPC delivers the full Data Intelligence Platform, including catalog, lineage, glossary, and AI-powered governance. It runs on networking, compute, and metadata storage provisioned for one organization, so governance metadata stays off shared infrastructure.
Impact: Regulated enterprises get full governance on isolated infrastructure.
Connectors add wider coverage and deeper lineage. Snowflake Bulk Retrieval Mode refreshes metadata in fewer passes at lower compute cost. Strategy V2 catalogs reports and traces their lineage to source systems, down to field level for Strategy datasets. BigQuery V2 adds table-level and field-level lineage. The new IBM DataStage connector catalogs jobs with table-level and field-level lineage.
Impact: Lineage reaches reports, BigQuery columns, and DataStage pipelines. Metadata refreshes on large Snowflake estates cost less to run.
MCP Server 1.1.0 lets any MCP-compatible AI assistant write descriptions, summaries, and custom properties into the catalog. Write tools prompt the AI client for confirmation first, and every change lands in the audit trail. Agents can also read lineage, audit history, pending suggestions, and contacts, with OAuth support included.
Impact: Documentation gaps close in the same conversation they’re spotted in. Every AI-written change stays confirmable and traceable.
Databases: Sovereignty
VectorAI DB is now available on AWS Marketplace. Teams buy through their existing AWS agreement and can draw down committed spend under an Enterprise Discount Program (EDP).
VectorAI DB also adds IVF indexing alongside HNSW, so customers can choose the recall, latency, and memory tradeoff that fits each workload. Quantization compresses each vector so the index fits the memory you have.
Impact: AI retrieval runs where your data lives, from cloud nodes to on-prem servers to edge devices.
FastObjects 15.0 brings three updates. Larger index keys let composite and long-string attributes take part in indexed lookups, so queries stop falling back to scans. A new hot index cache keeps frequently accessed index pages in memory and reduces contention under high concurrency. A new AES-GCM encryption option contributes to EU Cyber Resilience Act work for authenticated client connections.
Impact: Lookup-heavy applications keep predictable latency, and teams modernize without re-platforming.
NoSQL 11.1 adds logstream for disaster recovery. A standby database stays current over a live network connection, so recovery becomes a single promotion in place of a multi-step restore. Setup needs a standby host with network connectivity to the primary and roll-forward archiving enabled.
Impact: DBAs recover faster with fewer manual steps during an incident.
MCP Server 1.1 for Databases lets AI agents perform governed insert, update, delete, and merge operations. Each write authenticates through your identity provider, runs under the requesting user’s database identity, and passes human-in-the-loop approval. Teams can wrap complex ACID transactions using custom extensions. The server ships as a container image on Docker Hub and works with any MCP-compatible client.
Impact: Agents move from read-only Q&A to governed transactions, and security teams keep their approval gates.
Business Xchange (BX) adds 360 purchase order visibility with AI Analyst on BX data. Ask in plain language for a purchase order’s invoices, payments, and status, and get the full picture in one answer. BX also moves its platform to Kubernetes on Azure.
Impact: Finance and procurement teams reconcile purchase orders to payments in plain language without exporting data or building pivot tables.
Business Intelligence: Efficiency
Jaspersoft is now tested and certified to read Actian Zen, Ingres, and HCL Informix® databases out of the box, using the JDBC drivers you already have. Build pixel-perfect reports, interactive dashboards, and embedded analytics on the data you already hold. Jaspersoft deploys on-premises, in the cloud, or in hybrid environments.
Impact: Actian Zen, Ingres, and HCL Informix® customers get certified reporting on live operational data.
Take Action
- Explore Dedicated VPC and AI Analyst self-hosting with a Solutions Engineer. [Talk to a Solutions Architect]
- See Data Steward Agent, Data Contract Builder, and Lifecycle Policies in action. [Book a Personalized Demo]
- Get a custom demo of VectorAI DB. [Request Custom Demo]
- Explore the Jaspersoft commercial edition with a 30-day free trial [Try Jaspersoft 10.0]
Informix® is a trademark of IBM Corporation in at least one jurisdiction and is used under license.