←
Collate454 posts
Collate avatar

Collate

@CollateData

The AI-for-Data platform: agents that automate data management & AI analytics business users can trust. Built on OpenMetadata, the open context platform for AI.

◷ Joined May 2021▧ 369 Photos and videos
Collate avatar
Collate@CollateData·Aug 25

A valid-looking answer and a reliable answer from AI are not the same thing. Only when your LLMs are grounded in trusted context can you get trusted answers. Today @CollateData announces Collate 2.0, the AI for Data platform that provides agentic automation and orchestration to build enterprise context and put it to work. It carries the context from the question to the answer, into recurring workflows, through data-access decisions, out to the agents your engineers build, and into governance of the AI itself. And it is built on OpenMetadata, the Open Context Layer for AI agents, to ensure that this remains open and interoperable with all your AI and data systems. Select release highlights in Collate 2.0: - Context Layer - Context Center + Memory + Persona-Based Context Curation + Context API/MCP Tooling - Collate AI Analytics - Use natural language to build trusted analytics and governed dashboards with verifiable reasoning - AI Studio - Customize pre-built Collate agents, or build your own custom agents, with a powerful UI-based builder - Data Marketplace + Access Requests - Review Domains and Data Products for trusted data assets and quickly request access - AI Governance Studio (Release Preview) - Register, govern, and audit every LLM, MCP server, and agent Check out the Collate 2.0 release blog for all the details: buff.ly/Y4c5EDM #AIforData #OpenContextLayer #AIAgents #DataEngineering #DataGovernance #OpenMetadata

Post attachment
Collate avatar
Collate@CollateData·8h

Collate 2.1 in two and a half minutes. Ask an AI agent for last quarter's revenue, and it may find several tables with a revenue column, with no way to tell which one your business uses. Collate 2.1 gives agents that business context. The video walks through what shipped: — Ontology Studio: create, edit, and query your business ontology, bound to the data behind it. Start from FIBO or HL7 FHIR, or import your own — Memory Agent: learns from your documents and from the corrections your team makes in chat — Metrics: semantic and hierarchical metrics from Snowflake, Databricks, Looker, Power BI, and Tableau, plus AI Analytics on your own CSV and Excel files — Knowledge graph reasoning and approval workflows: agents find related metadata in fewer tool calls, and assets show only approved changes — Audit Hub (beta): a daily compliance score for your whole data estate, against standards such as BCBS 239 Collate 2.1 is built on the open-source @open_metadata project and on open standards such as OWL and SKOS, so your context stays yours. Watch: buff.ly/nZlIDmm #OpenMetadata #AIAgents #Ontology #DataGovernance

Collate avatar
Collate@CollateData·Nov 17

Many data tools now say they support ontologies. Most just give you a list of terms, but what an AI agent needs is a model of how your business fits together. That's what Collate's new Ontology Studio delivers. Create, edit, and query your business ontology in one place: — Model business concepts with their own properties and rules — Bind each concept to the tables and dashboards that implement it — Serve the model to agents through MCP, so one query returns a concept, its relationships, and its data — Start from industry ontologies like FIBO, HL7 FHIR, and GS1, then extend them for your business — Release versioning: publish a version for agents while your team keeps working on the next draft Built on open standards, so it stays yours when LLMs and vendors change. Read the blog: getcollate.io/blog/ontology-… #Ontology #KnowledgeGraph #AIAgents #OpenMetadata #ContextEngineering

Post attachment
Collate avatar
Collate@CollateData·Nov 17

AI agents can read your schema, but they cannot infer how your business works. Today, @CollateData announces Ontology Studio. It lets you document your business processes in an open-standard ontology you own, map it to the data that implements it, and make that context available to every AI agent that uses the data. — Model your business: concepts, typed properties, and named, directional relationships, with native support for OWL and SKOS — Keep it open and portable: import or export versioned ontologies in OWL, Turtle, RDF/XML, and JSON-LD — Start with ready-to-use ontologies for FIBO, HL7 FHIR, GS1 and EPCIS, ISA-95, and HR Open Standards As our CTO Sriharsha Chintalapani (@d3fmacro) puts it: "Ontology Studio keeps the ontology open and mapped to the data that implements it, so every agent a company runs works from the same definitions and business meaning that its people use." Ontology Studio is available in Collate 2.1. Read the announcement: getcollate.io/blog/announcin… #Ontology #AIAgents #DataGovernance #OpenMetadata #ContextEngineering

Post attachment
Collate avatar
Collate@CollateData·Nov 17

Ask an AI agent what revenue was last quarter, and it finds several tables in your data warehouse with a revenue column. It picks one without knowing which one your business actually uses. The problem is that nobody ever wrote down how the business works in a form your AI agents can understand. Our new Collate 2.1 release gives AI agents the ontology and context they need to connect your data to how your organization runs. Highlights: — Ontology Studio: create, edit, and query your business ontology, and connect it to its underlying data — Knowledge graph reasoning: Collate agents use the knowledge graph alongside vector search to retrieve related metadata with fewer tool calls — Memory Agent: automatically extracts memories from conversations and documents, then reconciles conflicts — Approval workflows: changes stay staged until an approver signs off, so assets show only approved metadata — Audit Hub (beta): a centralized governance dashboard that scores your entire data estate against a compliance standard such as BCBS 239, refreshed daily — AI Analytics: new advanced visualizations, plus analysis of your own CSV and Excel files Build your ontology on open W3C standards, including OWL and SKOS, so you can easily import/export or start from ready-to-use industry ontologies. Collate 2.1 is built on the open-source @open_metadata project to ensure your context stays yours. Read the announcement: getcollate.io/blog/announcin… #AIAgents #Ontology #DataGovernance #OpenMetadata #ContextEngineering

Post attachment
Collate avatar
Collate@CollateData·Nov 5

Can you show a regulator where a report number came from? Our guide covers which regulations call for lineage (BCBS 239, SOX, GDPR, Solvency II), what to look for in a tool, and how seven platforms compare. buff.ly/f0pSY1D #DataLineage #DataGovernance #OpenMetadata

Collate avatar
Collate@CollateData·Oct 29

In this episode of Ask the Experts, Nick Doyle and @progrockrec pick up from their previous conversation, discussing AI and a top-down vs. bottom-up approach. You need to have your context together for AI to make sense of your data accurately. Collate helps you do this in an agnostic way; we support whatever you're using, and we empower you to build the context you need to get the right answers. This discussion includes insights you won't want to miss. 🎥Watch here👉 buff.ly/rSqFItD #OpenMetadata #DataGovernance #DataDiscovery #aisemantics #Collate #contextinsights #DataEngineering #TechInterview

Collate avatar
Collate@CollateData·Oct 19

"If the foundations aren't there, every model will be confidently wrong." Live panel from @BigDataLDN: our CEO Suresh Srinivas, @MikeFerguson1, and Angelita Frozza Sanches (@Scout24) share experience from 1 year in production with AI agents. 🎥 youtube.com/watch?v=fmekRwap… #AIforData #ContextLayer #BigDataLDN #OpenMetadata