Data source connector builder Airbyte announced Airbyte Enterprise Flex, providing customers with data sovereignty and full control over their data, enabling accurate and secure data for AI and analytics. It says Flex is the industry’s first data movement platform built on one codebase, delivering the Airbyte feature set while giving customers the freedom to deploy wherever they choose – within days – in an on-premises datacenter, cloud, multi-cloud, or hybrid cloud. Airbyte also announced Data activation (often referred to as reverse ETL), which allows customers to move their data back into business-critical applications for AI and analytics use cases. Data Activation is available on all Airbyte plans for Hubspot and Customer.io destinations with plans to roll out to some 200 application destinations. Lastly it announced improved platform performance speeds – by as much as five to ten times. That translates to similar improvements in performance speeds for Airbyte supported connectors.
…
William Blair analyst Jason Ader tells subscribers that cloud and backup storage supplier Backblaze is in an ongoing business turnaround and led by a fresh management team outside of co-founder and CEO Gleb Budman. Management’s confidence in the business outlook stems from:
- The company’s competitive differentiation versus hyperscaler and neo-cloud storage offerings (less expensive, more efficient, and flexible storage)
- B2 Cloud’s growth reacceleration (driven by momentum with AI customers, increased cross-sell/upsell opportunities), which is projected to exit fiscal 2025 at 30-plus percent growth
- Major go-to-market changes (establishment of direct sales teams, focus on upskilling), which have already begun to bear fruit
- Secular AI/data tailwinds, which magnify the need for available storage at scale
- Improving profitability, with management prioritizing margin upside and adjusted free cash flow generation as core drivers here
…
Connector company CData Software announced Connect AI, the first managed Model Context Protocol (MCP) platform that integrates AI assistants, agent orchestration platforms, AI workflow automation, and embedded AI applications with more than 300 enterprise data sources. Connect AI has governed, in-place access to enterprise data, and, preserves data semantics and relationships, giving AI complete understanding of the context. It says Connect AI takes the same enterprise-grade connectivity technology already embedded by top technology companies including Palantir, SAP, Salesforce Data Cloud, and Google Cloud into their offerings, and reimagines it specifically for AI workloads with real-time semantic integration capabilities.
…
Cloudera has announced the launch of Cloudera Iceberg REST Catalog and Lakehouse Optimizer, and an integration with Dell ObjectScale to deliver a next-generation Private AI platform. The Iceberg REST Catalog enables secure, zero-copy data sharing and unified governance across any cloud or datacenter. With the Lakehouse Optimizer, enterprises gain automated cost savings and performance improvements for Apache Iceberg tables. Cloudera compute engines can run directly on ObjectScale storage, making it easier to manage both structured and unstructured data in one place with clear governance, security, and scalability.
…
Commvault and BeyondTrust have a strategic integration, bringing BeyondTrust’s Password Safe privileged access management (PAM) solution into Commvault Cloud, enhancing identity security.
…
Commvault research reveals that the rise of AI, tightening cross-border data regulations, and persistent cybersecurity threats are the top three external factors putting pressure on businesses to improve trust, with 97 percent of UK business leaders believing that introducing a chief trust officer role is necessary. This survey was conducted independently and exclusively for Commvault by Censuswide. It reveals the views of 1,000 UK business leaders, from companies with revenue of over £100 million.
…
Private equity biz Haveli Investments has acquired NoSQL database supplier Couchbase for $1.5 billion in cash four years after went public and since under-performed. Couchbase will now be a privately held company, and stockholders are entitled to receive $24.50 per share of Couchbase common stock owned immediately prior to closing. Sumit Pande, Senior Managing Director at Haveli Investments, said: “The combination of Couchbase’s strong product leadership with Haveli’s expertise in scaling enterprise software organizations, positions us well to expand market leadership while continuing to meet the performance and scalability demands of customers.”
…
Databricks and OpenAI announced the launch of a $100 million partnership that brings OpenAI models to all Databricks’ 20,000-plus customers. This means organizations can now bring OpenAI models directly to their enterprise data, benefit from access to high-capacity processing across the latest OpenAI models, and build production-ready AI agents by measuring the accuracy of models like GPT-5 and gpt-oss with task-specific evaluation and LLM judges. The $100 million number? It’s a minimum spending commitment from Databricks to OpenAI over the multi-year period of the deal.
If actual revenue exceeds $100 million, OpenAI receives more; if it falls short, Databricks covers the full minimum. This structure mirrors Databricks’ earlier $100 million deal with Anthropic and provides OpenAI with predictable income amid its datacenter expansions. It supports digital twins and autonomous systems, with applications already seen in Hitachi Rail’s HMAX.
…
Data integration supplier Fivetran is reportedly in talks to buy dbt Labs which provides an open source software framework to manage, transform, and model data in data warehouses.
…
Hitachi Vantara launched a global Hitachi AI Factory, built on Nvidia’s full-stack AI platform. The AI Factory has a distributed global infrastructure powered by Hitachi iQ with Nvidia HGX B200 systems featuring Nvidia Blackwell GPUs; Hitachi iQ M Series with Nvidia RTX PRO 6000 Server Edition GPUs; and Spectrum-X Ethernet networking platform. It enables seamless collaboration across the US, EMEA, and Japan, and provides acceleration of Lumada 3.0, Hitachi’s data-driven transformation model combining IT, OT, and AI to solve real-world challenges. (Lumada is Hitachi’s operating model that helps enterprises solve business and societal problems through co-created digital transformation.)
The Hitachi AI Factory is strategically distributed across the United States, EMEA, and Japan ensuring that Hitachi’s engineers can collaborate seamlessly and access powerful computing resources with low latency, no matter where they are.
…
Hitachi Vantara Federal announced Majed Saadi as the company’s CTO to lead Hitachi Vantara Federal’s technology strategy, driving IT modernization, strengthening enterprise platforms, and advancing the company’s position as a federal innovation leader. Saadi has held senior technology leadership roles across federal integrators, resellers, and consulting firms. As VP of Growth and Technology at Synergy Inc., he led market expansion, technology adoption, and federal contract growth. Previously, as VP at General Dynamics IT, he oversaw complex IT and mission support services for federal civilian agencies.
…
Lenovo announced SMB compute and storage products, which included “Business Protection in a Box: Safeguard critical data and workloads with Lenovo ThinkSystem SR650 V3 supporting up to 55 VMs; Lenovo ThinkSystem SR630 V3 + ThinkSystem Storage Arrays supporting up to 140 VMs.” Lenovo integrates Veeam to safeguard workloads against ransomware and failures, and with near-instant recovery, SMBs can restore critical operations in minutes, without needing a large IT staff.
…
Lenovo research drawn from its third Work Reborn report, Reinforcing the Modern Workplace, found 65 percent of IT leaders surveyed admitted their defenses are outdated and unable to withstand AI-enabled attacks, and just 31 percent feel confident defending against them. It says AI has changed the balance of power in cybersecurity. To keep up, organizations need intelligence that adapts as fast as the threats. That means fighting AI with AI. Lenovo says it’s leading with AI-native defenses designed to spot threats earlier, adapt in real time, and scale across the modern workplace.
…
OSI (Open Semantic Interchange) is an open-source initiative aiming to create a universal semantic data framework for all companies to standardise their fragmented data definitions with an open, vendor-neutral specification. In this starting era of AI agents, fragmented data definitions remain one of the biggest barriers to AI adoption, making a shared open standard essential to ensuring semantic consistency. OSI establishes a shared semantic standard so all AI, BI, and analytics tools can “speak the same language,” giving companies the flexibility to adopt best-of-breed technologies without losing consistency in metrics or business logic.
Without a shared semantic specification, data and AI teams often spend weeks reconciling conflicting definitions or duplicating work across platforms. By standardising how semantics are defined and exchanged, OSI hopes to ensure data is governed, consistent, and context-rich, enabling more accurate, trustworthy AI insights and faster adoption.
The OSI was set up by ThoughtSpot and Snowflake alongside Alation, Atlan, BlackRock, Blue Yonder, Cube, dbt Labs, Elementum AI, Honeydew, Mistral AI, Omni, Relational AI, Salesforce, Select Star, Sigma, and Tableau.
…
PostgreSQL has released v18 of its eponymous product. EnterpriseDB (EDB) contributed over 30 patches, more than ever before, helping to add 200 new features, including OAuth Authentication, SQL Standards Improvements and Performance Optimizer Enhancements, that make it easier for organizations to run secure, high-performance and portable applications across hybrid environments.
…
Peter Thiel’s Founders Fund-backed startup Sentient has launched its open AGI network, the first truly composable AI ecosystem where agents, models, and data sources work together. It’s accessible through Sentient Chat. Sentient’s mission is to ensure that Artificial General Intelligence (AGI) is open-source and not controlled by any single entity. While OpenAI and Anthropic operate closed ecosystems that limit what developers can build, and platforms like AWS’s AI agent marketplace and Hugging Face offer discovery of individual models without enabling coordination between them, Sentient orchestrates 40+ specialized agents into coordinated workflows capable of generating outputs like comprehensive investment reports that combine pricing, research, and market data in real-time.
At launch, Sentient’s platform features more than 40 specialized agents, 50 data sources, and 10-plus models. The agents, from both Web2 and Web3, include generative graphics engine Napkin, and fast-growing search startup Exa and Composio, a developer-focused platform that enables AI agents to seamlessly integrate with over 250 external tools and APIs.
…
Research biz TrendForce expects NAND flash prices to rise 5-10 percent in 4Q25, driven by spillover demand for QLC products. It says HDD shortages and longer lead times have prompted CSPs to quickly redirect storage demand toward QLC enterprise SSDs. SanDisk was the first to announce a 10 percent price increase, while Micron paused quotations due to pricing and capacity issues. QLC’s cost benefits have increased its adoption in SSDs. As generative AI drives higher demand for extensive data storage, suppliers are more focused on expanding QLC capacity.
…
VAST Data co-founder Jeff Denworth posted on X: “One of the silent stories of this week’s flurry of UK activity is the @VAST_Data involvement in these super-projects. CoreWeave, Microsoft/NScale, etc. There’s a whole lot of VAST’s AI Operating System headed to Great Britain!” VAST is at the #24 spot on the 2025 Forbes Cloud 100, “the definitive ranking of the top 100 private cloud companies in the world, published by Forbes in partnership with Bessemer Venture Partners.”
…
Veeam has promoted Kinda Baydoun as the new Lead for its EMEA Partner Organization. She has four years’ experience of leading partner engagement across the Middle East, English-speaking Africa, and Eastern Europe.
…
Mainframe data access supplier VirtualZ makes data available to LLMs and agents in four ways:
- Lozen – real-time, read-write access to mainframe data without replication, enabling governed AI/analytics directly against source systems.
- PropelZ – no-code ELT for landing mainframe data into JDBC targets (Snowflake, Databricks, SQL Server, etc.) with policy-driven control (e.g., column/row exclusion).
- Zaac – “instant mainframe storage” that bridges cloud object storage and z/OS for cost-efficient modernization.
- FlowZ – lightweight pipelines and backup/restore workflows that are getting traction with EU DORA-focused teams.
…
ReRAM developer Weebit Nano has joined the EDGE AI FOUNDATION as a Strategic Partner. The foundation is a global hub that unites industry leaders and researchers to drive innovation, solve global challenges, and democratize edge AI technologies. Weebit says it brings its low-power, high-performance Resistive RAM (ReRAM or RRAM) technology to this dynamic community. For advanced edge AI chips, Weebit ReRAM provides the dense on-chip non-volatile memory needed to store weights for artificial Neural Networks (NNs) with ultra-low power consumption that is critical for edge devices. It also scales to the smaller process geometries used in the fabrication of advanced AI SoCs.
