Latest AI and tech news
Making AI agents trustworthy starts with domain-scoped context and evals. Lessons from 5 months of beta as DataHub Cloud Context goes GA.
Context graph vs ontology: an ontology defines meaning, a context graph keeps it true. Why each needs the other and how to start small.
A unified context layer is one governed context graph for every person and agent. What unified means, and what it doesn't.
Data lifecycle management now runs as a loop, not a line. See the six stages, why AI changed them, and what has to travel with your data.
Data lineage for AI has two jobs: routing agents before they answer and tracing answers after. Learn why accuracy now matters most.
A context layer for AI agents has to be shared, stay current, and prove it works. Here are the three tests yours should pass.
AI agent context window management starts before the window. See why agents burn tokens on discovery, and how resolved context fixes it.
Data vs metadata: the line moves depending on what you're asking. A warehouse-level guide to the difference and why it matters.
What is metadata? Learn the definition, the main types of metadata, real examples from data teams, and why it matters for AI.
Data lineage impact analysis answers what depends on this. Here's how to run one, how far to trace, and what to do with the result.
DataHub Cloud 2.3 brings Context to public beta, with evals, agent feedback, and metrics in lineage so you can trust what your agents say.
CONTEXT 2026 is a free context management summit on November 4. See how teams build, govern, and monitor context for AI agents.
August 2026 town hall recap: the DataHub Metrics Catalog ships, Trustpilot governs 500,000+ assets, and Otto enters beta.
Six data lineage best practices, including the two most teams skip: giving lineage context and delivering it where people already work.
Data lineage tracking is only as good as its coverage. Why capture that stops at one platform's boundary hides what breaks downstream.
AI agent context is more than what fits in the window. See the four layers agents need, and how each fails silently in production.
“Welcome to DataHub! Hey, we’re working on this Otto agent, can you help us launch it?” That was…
Context engineering for AI agents works until you scale it. What enterprise programs need beyond the context window, and how to build it.
What a business context layer holds, why a glossary isn't one, and how Miro took agent accuracy from under 40% to over 90%.
An enterprise data catalog has to handle scale, compliance, and AI readiness. Learn what that requires and when you actually need one.
Data lineage in a data mesh is what keeps domain autonomy from turning into silos. See how cross-domain lineage holds a mesh together.
How DataHub and SecuPi pair trusted context with runtime enforcement so AI agent governance follows the end user.
Seven projects shared $20,500 building agents on DataHub's context graph. See what they built and how they used real lineage, schemas, and write-back.
DataHub Cloud v2.2 opens the Context Platform to Public Beta, ships custom AI Agents in private beta, and introduces an Ontology Explorer and Data Product Marketplace.
How DataHub integrates with modern tech stacks: what it ingests from 150+ sources and how that context reaches your agents and BI tools.
A unified data platform brings data discovery and governance together. See the benefits, the proof, and what fragmentation really costs.
Miro's text-to-SQL climb past 90%, Just Eat Takeaway running DataHub self-hosted at scale, and a look at the Metrics Catalog. July 2026 Town Hall recap.
Interoperable context isn't whether context exists, but whether it travels. Why platform-native context layers aren't enough on their own.
AI-ready data is more than clean pipelines. See the 5 pillars, the 2026 readiness gap, and what AI agents actually need.
Context preparation is important because institutional knowledge is unevenly held, a human gap agents made impossible to ignore.
Data interoperability isn't a connector count. What it means, why context gets trapped in an agentic stack, and how to evaluate it.
Data governance is important because AI acts on what your metadata says. What it delivers, what it costs to skip, and what to fix first.
The DataHub community has always been global. Nearly half of users are outside the United States, and the…
A context platform is only as useful as the context it can reach. With v2.1, DataHub Cloud extends that reach: agents and the services…
Every data team building conversational analytics eventually runs into the same problem: the agent gives you an answer, and you have no idea whether…
The five modern data catalog features that separate static inventories from AI-agent-ready infrastructure. Evaluate what your catalog has.
Databricks Genie One explained: what it is, how Genie Ontology powers it, and where it fits your stack.
What is Snowflake Cortex Sense?Snowflake Cortex Sense is a runtime context layer, announced at Snowflake Summit in June 2026, that automatically learns business definitions,…
A year ago, the open question was whether a data catalog could feed an AI agent at all.…
Compare the best context management platforms for AI agents in 2026: enterprise context platforms, data catalogs, and open-source options.
Azure data lineage shouldn't stop at the Microsoft edge. See how DataHub adds column-level depth and unifies it across your whole stack.
Lineage in ETL only works when it meets three conditions: column-level, cross-platform, and captured at runtime. Here's why.