The decision reflects a broader push toward agentic, AI-powered observability. By removing financial and operational barriers, Edge Delta aims to simplify adoption of its AI Teammates—autonomous agents designed to monitor, analyze, and act on system data in real time. These agents go beyond traditional observability tools by proactively identifying anomalies, correlating signals, and initiating investigations without human intervention.

Unlike conventional pipeline tools that require extensive engineering effort and primarily focus on data movement, Edge Delta’s offering combines data routing with intelligent automation. Its pipelines act as a preprocessing layer, refining logs, metrics, and traces before they are analyzed, ensuring higher-quality insights and reduced noise.

Also Read: Arango Unveils Contextual Data Platform 4.0 to Accelerate Enterprise AI Deployment

“The telemetry pipeline market is crowded, manual, and largely unintelligent. Cribl, OpenTelemetry, Bindplane, Databahn, Fluent, and Vector require significant engineering effort to deploy and maintain, yet at the end of the day still deliver little more than moving data. Today, Edge Delta Telemetry Pipelines is now free at any scale. No throughput limits. No per-GB fees. Any organization can route, transform, and control unlimited telemetry volume at no cost. This is a deliberate strategic decision, not a promotional one. The demand signal around Edge Delta AI Teammates has been clear: enterprises are ready for observability that thinks and acts, not just collects and displays. Making pipelines free removes the last barrier to experiencing that firsthand. Operations teams have watched AI transform developer workflows for years now. Edge Delta AI Teammates bring that same capability to the people running production infrastructure. The path is now frictionless.” — Ozan Unlu, Founder & CEO, Edge Delta

This kind of framework allows organizations to deal with gigabytes or even petabytes of telemetry data at zero costs. In addition, this trend conforms to the current industry trend of automating the management of operational processes using AI tools that will minimize human intervention in incident resolution.

The combination of Edge Delta’s free data pipeline services and its innovative AI-powered observability layer will transform the way engineering teams operate in complex and highly dynamic environments.

The post Edge Delta Eliminates Data Throughput Costs to Accelerate AI-Driven Observability Adoption appeared first on ITDigest.

While enterprises are looking to take AI from a proof-of-concept phase into production, the pain points associated with fragmented data systems and complex integration scenarios have become more pronounced. Most traditional methods seek to rebuild relationships between data sets at the inference layer, resulting in non-consistent results and a lack of transparency. Arango’s latest platform addresses this by embedding contextual modeling directly into the data layer, allowing enterprises to maintain a continuously updated and governed data foundation.

The Agentic AI Suite is a major part of the release. It comprises over 20 built-in AI services as well as exclusive tools such as AutoGraph, AutoRAG, and Arango Ada. These tools automate essential tasks such as data ingestion, contextual modeling, retrieval optimization, and workflow orchestration, therefore greatly decreasing the engineering work needed to go from development to production.

Also Read: Unstructured and Teradata Partner to Scale AI-Ready Data

To give an example AutoGraph is responsible for automatically arranging structured and unstructured data into interconnected knowledge graphs, which allows AI systems to comprehend the relations between business entities and events. Meanwhile, AutoRAG enhances retrieval strategies by combining graph-based, vector, and hybrid search techniques, ensuring more accurate and context-aware outputs. Arango Ada further simplifies development by allowing users to interact with complex data systems through natural language queries.

The platform also offers a flexible deployment option of Bring Your Own Code/Container (BYOC) model. Thus, organizations can integrate their preferred AI models while still having control over security, governance, and compliance requirements.

Being highly scalable, the platform can be deployed on cloud, on-premises, hybrid, and air-gapped environments which make it suitable for regulated industries and very large-scale enterprise operations. By offering a unified contextual data foundation, Arango intends to assist organizations in developing AI systems that are not only scalable but also explainable, traceable, and in line with business conditions.

The post Arango Unveils Contextual Data Platform 4.0 to Accelerate Enterprise AI Deployment appeared first on ITDigest.

Teradata has embedded Unstructured’s data processing platform natively inside Teradata Enterprise Vector Store, giving customers a secure path to transform documents, images, video, and audio into AI-ready data without external tools or pipelines

The post Unstructured and Teradata Partner to Scale AI-Ready Data appeared first on ITDigest.

The new Streaming Agents features enable AI agents to communicate, coordinate, and take actions on real-time data streams in real time, effectively removing the barriers that have historically existed between analytics, operational data, and automated decision-making processes. By providing continuous context to AI agents and allowing them to interact with each other and other systems, the Confluent solution is intended to enable a system of collaborative, context-aware AI workflows that can identify patterns, prevent problems, and take actions at speeds never before possible – effectively translating raw data into actionable enterprise intelligence.

According to Confluent’s Head of AI, Sean Falconer, businesses must evolve beyond batch-oriented analytics and adopt ecosystems where AI agents work in concert and learn from fresh signals – not just historical snapshots – to stay competitive.

Also Read: Databricks’ DLT-META Brings Order to Big Data Pipelines 

The New Capabilities Explained

Support for Agent2Agent (A2A) Protocol

A core part of this announcement is Confluent’s support for A2A, an open protocol that enables multiple AI agents to exchange information, coordinate tasks, and share context without custom engineering between systems. In practical terms, this means an AI agent in a CRM might trigger actions for another agent handling supply chain events — all in real time based on business signals flowing through Confluent’s data streams.

This feature enables businesses to create a network of interlinked agents instead of having disconnected AI tools, which in turn reduces redundancy and speeds up the process of automated decision-making. This feature also enables businesses to have auditability and governance through streaming observability.

Multivariate Anomaly Detection

ML can make another big difference here by implementing Multivariate Anomaly Detection. Instead of looking at individual metrics separately, like CPU usage, memory, and latency, it looks at them all together to detect anomalies. Hence, businesses can recognize new trends that without this technique, would have been overlooked, so they can fix problems even before they deteriorate.

For enterprise IT, it basically means the system will notify them ahead of time if it is going to run into a performance issue, there is a security breach or customers are behaving differently which none of these will be guesses but data, driven and based on real, time information rather than batch processing.

Implications for the B2B and Business Data Industry

Confluent’s expanded intelligence capabilities come at a time when digital transformation and AI adoption are top priorities for enterprises across industries. In the broader B2B and business data ecosystem, this development is significant for several reasons:

Bridging Real-Time Data and AI for Competitive Insight

Many B2B organizations have struggled to implement AI because traditional data systems are based on batch processing, where the data is updated periodically but not fresh. With the help of Confluent, organizations can now build systems that respond to real-time events as they occur, whether it is identifying a supply chain disruption or making a customer offer in real time.

The concept of continuous context helps organizations create more responsive systems and make decisions faster. It also helps organizations gain insights that were not visible until after the event.

Operationalizing AI Across the Enterprise

AI adoption performs best when systems are not siloed and can share a unified data context. Confluent’s use of A2A protocols and real-time streams means that AI agents in different functions — sales, logistics, finance, customer service — can coordinate actions based on the same data foundation.

For B2B firms, this promises better operational alignment, improved workflow automation, and more accurate predictions — whether optimizing fleet logistics in manufacturing or automating credit risk evaluation in financial services.

Reducing Data Silos and Improving Governance

One of the biggest challenges facing enterprise-level AI initiatives is that of fragmented data environments. Without the ability to access real-time business context, AI models are forced to make decisions based on outdated or incomplete data, resulting in suboptimal performance and a lack of trust in the system. Confluent’s platform provides a governed data stream that provides AI systems with clean and trustworthy data in real time, a necessity for scaling AI in large enterprises.

Faster Time-to-Value and Lower Operational Risk

Real-time anomaly detection and agent coordination make it easier to conduct experiments for AI projects. Teams will not have to wait for weeks to conduct batch analytics. Instead, they can test, iterate, and deploy AI processes with confidence that they are working on new and accurate signals. This will minimize risks and speed up the deployment of intelligent applications that affect revenue, customer satisfaction, and efficiency.

Wider Business Impacts and Future Trends

Confluent’s vision fits into a larger trend in the industry towards event-driven architecture, where data is not only stored but also flows constantly, driving real-time responses and predictive analytics. According to larger industry trends, the future of enterprise AI will demand strong infrastructure that feeds AI models with constant and trustworthy data streams, rather than static points in time.

With the growing complexity of data systems in the enterprise and the need for AI adoption across business functions, technologies that integrate real-time streaming and AI, such as Confluent’s enhanced Intelligent offerings, may form the basis of a new generation of business infrastructure.

From B2B customer experiences to risk management and supply chain optimization, the ability to convert data into actionable insights in milliseconds, rather than hours, may well be the key differentiator.

Conclusion

The recent improvements to the Intelligence platform by Confluent, such as the ability to work with collaborative AI agent ecosystems and anomaly detection, represent a turning point in the way that organizations can operationalize real-time data and enterprise AI. With the ability to provide systems that respond to real-time signals and act in concert across departments, Confluent is enabling businesses to overcome the traditional silos that exist between data, AI, and automation.

In the B2B and business data industry, this is a technological advancement as well as a strategic opportunity.

The post Confluent Intelligence Expands Real-Time Business Data to Enterprise AI appeared first on ITDigest.

Also Read: AWS Enables Durable, Production-Ready AI Agents with LangGraph and Amazon DynamoDB 

The jina-embeddings-v5-text models are available as open-weight models through Hugging Face for self-hosted implementations using frameworks such as vLLM, llama.cpp, or MLX, while also being integrated into Elastic’s GPU-powered Elastic Inference Service (EIS). Through EIS, enterprises can run inference workloads without managing complex infrastructure, allowing them to deploy advanced AI-powered search and retrieval capabilities more easily across cloud and on-premises environments. According to Steve Kearns, general manager, Search at Elastic, “Vector search, RAG, and AI agents depend on high-quality retrieval,” said Steve Kearns, general manager, Search, Elastic. “With the addition of the Jina v5’s multilingual embeddings, Elasticsearch continues to be the platform of choice for end-to-end context engineering.” In general, the launch is a continuation of Elastic’s strategy to combine search technology with AI capabilities, which will enable enterprises to develop more intelligent applications that can turn massive amounts of data into valuable insights, including workloads like generative AI and retrieval-augmented generation.

Read More: Elastic Introduces Best-in-Class Embedding Models for High Performance Semantic Search

The post Elastic Launches High-Performance Multilingual Embedding Models for Semantic Search appeared first on ITDigest.

Also Read: Databricks’ DLT-META Brings Order to Big Data Pipelines 

Developers are also shown how-to guides for creating a DynamoDB table as a prerequisite along with an S3 bucket, example code for creating a DynamoDB Saver using a workflow involving a LangGraph, as well as how-to guides for getting checkpoints for either debugging or auditing tasks. The article emphasizes that moving from prototype to production can be as simple as switching from memory-based checkpointing to the DynamoDBSaver to gain persistence, durability, and scalability. “By integrating DynamoDBSaver into your LangGraph applications, you can gain durability, scalability, and the ability to resume complex workflows from a specific point in time.” AWS also recommends considering Amazon Bedrock AgentCore Runtime for fully managed operational environments to handle scaling and monitoring while developers focus on agent logic.

This integrated approach with DynamoDB and LangGraph empowers enterprises to build robust AI workflows that maintain context, support auditability, and handle complex state across diverse production environments reliably and efficiently.

Read More: Build durable AI agents with LangGraph and Amazon DynamoDB

The post AWS Enables Durable, Production-Ready AI Agents with LangGraph and Amazon DynamoDB appeared first on ITDigest.

As data volumes, sources, and analytic complexity grow, the challenge for data teams isn’t just compute speed it’s consistency, governance, and maintainability. Traditional approaches demand extensive, manually written pipelines for each dataset, leading to siloed logic, duplicated effort, and uneven standards across teams. Databricks’ new framework aims to replace this ad hoc chaos with templated, metadata-centric pipeline generation that enables organizations to focus more on business logic and less on pipeline plumbing.

What Is DLT-META?

At its core, DLT-META is a metadata-driven metaprogramming framework that works with Spark Declarative Pipelines the successor to Delta Live Tables and part of the Databricks ecosystem for defining intent-based ETL. Rather than hand-coding a separate pipeline for every data source or table, teams define their logic, data sources, transformations, quality rules, and governance expectations in structured metadata (JSON/YAML). DLT-META then dynamically generates the underlying pipeline at runtime.

This approach centralizes rule definitions and pipeline logic into a single source of truth, reducing errors, improving repeatability, and enforcing standards organizationally. As the Databricks blog notes, metadata first means “pipeline behavior is derived from configuration, rather than repeated notebooks.”

Why This Matters to the Big Data Industry?

The Big Data landscape has long grappled with scale, velocity, and complexity. Tools such as Apache Spark provide the horsepower for distributed computation, while modern data stack components like Databricks’ Lakehouse unify storage and analytics. Yet as enterprises ingest hundreds or thousands of data feeds, the orchestration overhead can quickly become a bottleneck not in terms of processing power, but in engineering time and maintainability.

According to the Databricks post, manual pipelines struggle at scale because each new source adds “too many artifacts per source,” and schema changes trigger extensive rework across dozens of pipelines. DLT-META’s metadata-first approach directly addresses this by enabling a single template to handle many similar pipelines, reducing manual effort and ensuring consistency.

This fits broader trends in the Big Data ecosystem toward declarative paradigms where users describe what they want, and the engine handles how to execute it. Databricks recently open-sourced its declarative pipeline framework into the broader Apache Spark project, signaling a shift toward standardizing these patterns across platforms.

Also Read: Red Hat and NVIDIA Deepen Partnership to Accelerate Enterprise AI Adoption with Rack-Scale Innovation

Business Impacts: Faster Delivery and Lower Costs

Businesses that rely on data to make decisions whether retail for inventory optimization, finance for risk analytics, or healthcare for patient insights depend on reliable and timely data processing. DLT-META can help organizations:

• Accelerate time to production: Adding a new data source or business rule could go from weeks of engineering work to minutes of metadata configuration.
• Improve quality and governance: With consistent templates, logic drift across teams is reduced, enabling easier compliance with rules, lineage tracking, and audit requirements.
• Reduce maintenance overhead: Metadata-driven pipelines minimize repetitive code and reduce the burden of updates when schemas or business logic change.
• Enable broader team involvement: Domain experts can contribute through configuration rather than code, freeing specialized data engineers to focus on high-value tasks.

For organizations already using Databricks and Spark Declarative Pipelines, adopting DLT-META offers a practical path to scalability without sacrificing data quality or governance. And because Spark Declarative Pipelines which replaced the older Delta Live Tables (DLT) project are being open-sourced within Apache Spark, the benefits of this approach are poised to extend beyond a single vendor ecosystem.

Challenges and Considerations

However, adoption isn’t without caveats. DLT-META is currently a Databricks Labs project, meaning it is provided for exploration and lacks formal support or SLAs from Databricks itself. Organizations should factor this into their risk assessments, especially for critical production workloads.

Moreover, while metadata-driven automation sounds ideal, it also requires good metadata hygiene. Poorly defined metadata can propagate errors at scale just as quickly as manually coded pipelines, underscoring the importance of governance and validation practices when adopting the framework.

Looking Ahead

As data environments grow more complex and data teams struggle with both scale and velocity, innovations like DLT-META represent an important evolution in pipeline engineering. By leaning into metadata and declarative paradigms, the Big Data industry can shift from bespoke coding patterns toward repeatable, governed, and scalable processes.

For businesses, this isn’t just a technical improvement it’s an operational one. Faster onboarding, consistent governance, and reduced engineering toil can translate into faster insights, lower costs, and better alignment between data assets and business outcomes.

The post Databricks’ DLT-META Brings Order to Big Data Pipelines appeared first on ITDigest.

The Integrated Stack: Hardware Meets Open Source

The partnership guarantees Day 0 support for the Vera Rubin platform across Red Hat’s entire hybrid cloud portfolio. This means that when the Rubin architecture ships in the second half of 2026, the software ecosystem will already be optimized for its 3,600 GB/s NVLink speeds and HBM4 memory.

Key Integration Pillars:

• RHEL for NVIDIA: A specialized foundation optimized for the Vera CPU and BlueField-4 DPU, featuring native support for NVIDIA Confidential Computing to protect model weights and sensitive data-in-use.

• Red Hat OpenShift: Full integration with CUDA X libraries, enabling automated lifecycle management for distributed AI workloads on Kubernetes.

• Red Hat AI: Expanded support for distributed inference, allowing organizations to run NVIDIA’s latest open models (like the Nemotron family) across hybrid cloud environments with near-zero friction.

Also Read: NVIDIA Expands AI Customization with Unsloth Fine-Tuning on RTX PCs and DGX Spark

Why the Vera Rubin Architecture Changes the Game

The Vera Rubin platform is not just a chip upgrade; it is a fundamental shift to extreme co-design. By integrating six custom chips—including the Vera CPU and Rubin GPU—into a unified rack-scale supercomputer, NVIDIA is claiming a 10x reduction in inference token costs compared to previous generations.

Strategic Impact for Decision Makers

For the enterprise, this collaboration addresses three critical hurdles:

1. Time-to-Market: Day 0 support eliminates the “integration lag” that usually follows a new hardware launch.

2. Security at Scale: With hardware-rooted Confidential Computing now standard in the RHEL/Rubin stack, regulated industries (Healthcare/Finance) can finally move proprietary models to the public cloud without risking intellectual property.

3. TCO Optimization: Unified rack-scale designs reduce power, cooling, and management overhead, lowering the total cost of ownership for high-density AI infrastructure.

“To meet these tectonic shifts at launch, Red Hat and NVIDIA aim to provide Day 0 support for the latest NVIDIA architectures,” said Matt Hicks, CEO of Red Hat. “Together, we are fueling the next generation of enterprise AI through the power of open source.”

The post Red Hat and NVIDIA Deepen Partnership to Accelerate Enterprise AI Adoption with Rack-Scale Innovation appeared first on ITDigest.

The enhanced observability capabilities include four new event types that enable real-time information on export progress, status extended, and table-level notifications regarding long-running export jobs. The customers can observe key export progress metrics such as the number of tables that have been exported successfully, the number of tables that are pending export, and the amount of data that has been transferred.

With such improvements, users can now subscribe to export notifications using Amazon Simple Notification Service (SNS) to receive notifications and view the event data using the AWS Management Console, AWS Command Line Interface, or AWS SDKs. The observability improvement not only benefits operational transparency but also helps in speeding up troubleshooting and performance planning for data export activities.

Also Read: BigID Introduces “Agentic AI Remediation” to Accelerate Risk Reduction at Scale 

“With this new update, customers now have the ability to view their export job progress in greater detail. This also perfectly combines with the AWS services for monitoring and notification,” said an AWS spokesperson. “Additionally, this new feature from AWS will help organizations maximize their data pipeline operations.”

The export of database snapshots from Amazon RDS to S3 enables organisations to export database snapshot data from their databases in order to store this data in Apache Parquet format in S3 buckets, which can then be processed using various AWS services such as Amazon Athena, Glue, or Redshift Spectrum, while increased observability adds to this feature.

The post Amazon Web Services Enhances Observability for Amazon RDS Snapshot Exports to Amazon S3 appeared first on ITDigest.

Enterprises are increasingly spanning multiple data stores, from cloud object stores such as AWS S3 to collaboration tools like Microsoft SharePoint, Google Drive, OneDrive, and legacy on-prem file shares. Due to this, the volume and complexity of data activity have exploded. According to BigID, this has left many organizations blind to critical questions such as: Who accessed or modified data, when, and what data was involved.

Activity Explorer aims to bridge that visibility gap. It consolidates fragmented logs and audit trails into a single, searchable interface to let security teams track data activity across identities-human users, service accounts, automated workflows, and AI agents alike across resources like cloud, SaaS, and on-prem via operations in download, deletion, modification, sharing, and over time.

This promises not only faster, more accurate investigations, for instance “Who deleted that file?”, “Which sensitive files were downloaded during this time window?”, but also richer context including data sensitivity classification, permissions, ownership, and exposure risk. Mixed with compliance requirements under such regulations as GDPR, HIPAA, CCPA, and others, the new tool provides audit-ready activity logs for enterprises.
According to its leadership, BigID considers this as the next evolution of its platform: delivering DSPM, data intelligence, cloud data loss prevention (DLP), and now activity monitoring in a single, unified way.

Why “Activity Explorer” Matters for Big-Data and Enterprise Data Management

1. Bridging the Visibility Gap in Distributed, Hybrid Data Architectures

Few modern data architectures, especially in enterprises, are monolithic. Organizations now have data strewn across a multitude of clouds, SaaS apps, on-premise servers, and hybrid storage systems. This complexity is the reason why traditional logging and monitoring are grossly inadequate. As BigID argues, many of these legacy tools were built in a world where data was centralized and the identity models were simple.

Blind spots in who accessed what and from where are a major risk for big-data enterprises that could be processing petabytes of structured and unstructured data across on-prem Hadoop or data-warehouse clusters and cloud data lakes. Unauthorized access, accidental data exposure, or insider threats are hard to detect without unified visibility. Activity Explorer correlation of data sensitivity with actual usage is a major step forward in data governance for big-data environments.

2. Supporting Compliance, Risk Management, and Audit Readiness

As the volume of personal or sensitive data collected, stored, and processed by businesses increases, so do the demands of regulation and compliance. Be it GDPR, HIPAA, CCPA, or any other region-specific data-privacy law, it is not only important to prove that data is stored securely but also that access to it is tracked and auditable.

It provides a single audit trail across hybrid data sources-cloud, SaaS, and on-prem-enriched with data classification and ownership context that will help businesses produce compliance-ready evidence. This reduces legal, compliance, and reputational risk-a big benefit for enterprises in regulated industries such as healthcare, finance, and retail.

3. Improving incident response, insider-threat detection, and data breach containment

Large-scale data incidents often arise not just from external attacks, but due to insider threats, misconfigured permissions, or misuse by service accounts or automated agents. For enterprises using big data for analytics, AI, or machine learning, automated workflows and AI agents are common yet poorly monitored.

Activity Explorer monitors activity from human and non-human identities alike-service accounts, AI agents, automation-that gives security teams clarity into what happened, but also how and by whom. In the case of a breach or suspicious behavior, teams can rapidly trace the “blast radius”-which files, datasets, or repositories were touched-and contain exposure.

4. Enable Smarter Data Governance & Least-Privilege Enforcement

Above all, over-permissioned user accounts, sprawling access rights, and unmonitored usage of data build up over time as an organization grows into teams and departments with numerous sources of data. Activity Explorer applies identity and permission context to the real-world activity observed and monitors how data is actually being accessed and consumed. This information can fuel “least-privilege” efforts by removing unnecessary or risky permissions and reinforcing zero-trust policies.

Also Read: AWS Launches Kiro and MCP (Model Context Protocol) Tools for SQL Server Professionals – A Big Shift for Big Data Operations 

What this means for the greater big-data industry and businesses:

The introduction of Activity Explorer is indicative of the trend in the big-data and data-management world: from purely storing and processing volumes of data to governing, protecting, and making sense of data usage at scale. As generative AI, machine learning, and analytics adoption grows, data environments are becoming truly dynamic, distributed, and complex. With that complexity comes risk, and tools like Activity Explorer signal that security and compliance are becoming first-class citizens in big-data ecosystems.

But for businesses, especially those that are dealing with sensitive personal data, regulated industries, or complex data pipelines, this may raise the ante on data governance. We will likely see more people adopting unified DSPM, monitoring, and compliance platforms. This is better than using a mix of separate tools. Vendors providing big-data storage, analytics, AI insights, or data-lake services must enhance their audit and monitoring features. They might also consider partnering with companies like BigID.

As cloud migration grows, hybrid environments will become standard. This includes cloud, on-prem, SaaS, and AI workloads, even for large enterprises. Tracking data activity in a hybrid mix helps organizations gain better control and see risks more clearly. This will shape how companies design their data and AI systems. They will focus on built-in observability, compliance, and least-privilege access. The rise of tools like Activity Explorer shows that big data is maturing. It’s not just about size and speed anymore. Now, it’s about smart data, governance, and security. This is vital in a world where data is regulated, spread out, and used by AI.

The post BigID Unveils “Activity Explorer” – A New Era of Unified Data Monitoring appeared first on ITDigest.