The next chapter of CluedIn: building the foundation for Agentic Data Management

AI-assisted data management

Making AI practical inside real data operations, from lookup resolution and deduplication to rule creation, data quality prioritization and guided remediation.

Agent-to-agent interoperability

Preparing CluedIn to participate in wider AI and automation ecosystems, including support for safer, clearer and more efficient agent interactions.

Performance, observability and cost optimization

Improving the operational backbone of the platform so teams can monitor, troubleshoot, scale and optimize CluedIn with greater confidence.

Modernized data operations and activation

Strengthening workflows, streams, data contracts, export receipts and write-back patterns so trusted data can be governed, activated and measured downstream.

Agents Redesign

A richer agent experience designed to make agents easier to configure, monitor and apply across data quality, lookup resolution, deduplication, rules and operational workflows.

Self-Valuing Loop

Agents will be able to evaluate usefulness, confidence and business value, helping workflows improve over time and reducing unnecessary manual review.

RAG and vectorisation

Retrieval-augmented generation and vector-based representations will help agents work with relevant CluedIn context, improving recommendations, semantic matching and resolution workflows.

What this unlocks

• Faster creation of validation rules
• Generated transformation logic
• Recommended rule corrections
• Explanations of existing Power Fx expressions
• Safer rule drafts before user approval

The bigger shift

This is not AI as a black box. It is AI helping users create deterministic, inspectable and governable logic. That distinction matters. Enterprise data teams need speed, but they also need control.

CluedIn as an intelligent participant in the wider AI ecosystem

CluedIn will introduce an MCP layer designed for agent-to-agent interoperability. The goal is to help MCP-enabled systems understand how to work with CluedIn safely and efficiently.

• How to query and interact with CluedIn safely
• How to handle errors and exceptions
• How to reduce unnecessary token usage
• How to optimize requests for cost, accuracy and performance
• How to understand CluedIn’s capabilities, constraints and recommended interaction patterns

Processing Pipeline Update

Improvements to reliability, throughput and operational visibility, supporting larger data volumes, reducing bottlenecks and strengthening recovery from failed processing steps.

Jobs Monitor

A clearer view of active, queued, completed, failed and retried jobs, giving administrators a single place to understand what is running, what is blocked and what needs attention.

OpenTelemetry support

Improved observability across services, processing jobs, APIs and infrastructure, helping teams troubleshoot faster and analyze performance with greater confidence.

Lower cost of running

Continued focus on infrastructure efficiency, workload optimization, smarter processing behavior, improved storage patterns and better visibility into expensive operations.

Data Contracts

Stronger governance and operational management over data usage, access, movement and lifecycle, supporting enterprise controls around quality, retention, stewardship, compliance and downstream activation.

Streams V3

Improvements to stream mapping and projections, making streams more flexible, predictable and easier to use when preparing data for downstream systems.

Export Target receipts

Clearer confirmation of what was sent, where it was sent and whether the receiving target accepted or processed the payload.

Workflow Engine V3

Modernized workflow capabilities including write back, reverse ETL, custom data sources and custom rules to support more operational activation patterns.

Lookup AI Resolution

AI-assisted matching and recommendations to suggest likely matches, identify ambiguous values and reduce manual mapping effort.

Duplicate Clusters AI Loop

AI-assisted duplicate review to explain why records are clustered, identify risky merges, recommend next actions and continuously improve outcomes.

Rule Inspector

A clearer way to understand how rules execute, why outcomes occurred and how rule logic can be improved before it creates problems in production.

AI foundation and operational improvements

Agents Redesign, Self-Valuing Loop, Processing Pipeline Update, Lower Cost of Running and Performance UI/UX updates.

AI-assisted modeling and platform visibility

RAG, vectorisation, MCP, Jobs Monitor, Data Controls, Workflow Engine V3 and the Streams V3 foundation.

Observability, activation and rule workflows

Data Quality Metrics, agents writing Power Fx, OpenTelemetry, Blob Storage migration, Streams V3, Export Target receipts, Rule Debugger and write back / reverse ETL.

AI resolution and topology

Lookup AI Resolution, Duplicate Clusters AI Loop, Topology Viewer 2, Custom Rules and Workflow Engine V3 improvements.

Ongoing improvement

Performance UI/UX updates, lower cost of running, agent skills, observability improvements, UX polish and stability fixes.

For data leaders

A clearer way to move from fragmented data and manual governance to measurable, AI-ready data operations that can scale across the enterprise.

For data stewards

More guidance, better prioritization, smarter resolution and less time spent manually interpreting every quality issue, lookup problem or duplicate cluster.

For data engineers

Stronger observability, improved processing foundations, better stream handling, clearer export confirmation and more operational control.

For business users

A more guided experience that helps the business participate in data improvement without needing to understand every technical detail behind the platform.