Atlan

⏩ Aligning your data org with AI-era demands? Here's how roles are shifting: Data modelers → Context engineers Data architects → Knowledge graph builders AI engineers → AI evangelists DigiKey's CDAO Sridher Arumugham has been preparing his team for the future and they've already launched 70 AI projects with this foundation. 💬 His philosophy: While AI automates the manual work, data teams must focus on building the semantic layer that teaches AI what data actually means. 🎥  Watch Sridher talk about how AI is changing data teams at #ReGovern2025 https://lnkd.in/d4riYbVn

🤖How to build AI Analysts that actually work in production. Your AI guesses business meaning. And guessing breaks trust. Through Atlan AI Labs, Shubham Bhargav (Product & Engineering at Atlan) and team worked with some of our most AI-forward customers like Workday to figure out what it actually takes to build AI Analysts that teams trust in production. The result? A 5x increase in response accuracy.📈 What it took: Instead of dumping metadata into prompts, we built a structured context layer: → Rich, queryable metadata (not static docs) → Domain-specific definitions (not generic glossaries) → Continuous validation loops (not one-time setup) The AI Analysts that work in production aren't just connected to data. They're grounded in meaning. Here's a full guide ➡️ https://lnkd.in/df2KhUZ3

A CIO once told us about their AI rollout: "Models were easy. Data and context? Brutal" That’s the hidden reason most AI projects fail. It’s not the models. It’s the context gap: The missing bridge between your data and your business reality. In her #ReGovern2025 keynote, our Founder & Co-CEO Prukalpa ⚡ breaks down why this gap stalls AI pilots: ➡️ Context is scattered → AI can't find the tribal knowledge buried in Slack threads, CRMs, and SharePoint. Without it, models hallucinate. ➡️ Business meaning is missing → LLMs understand language, but not what "customer" or "revenue" means inside YOUR company. ➡️ Governance doesn't scale → Pilots are controlled environments. Production isn't. Permissions sprawl, regulations shift, and risk multiplies. Until teams close this gap, most AI will remain stuck in demo. Watch the full keynote to learn how to bridge this gap👇  https://lnkd.in/dN-N_3eD

Data trust starts with visibility. See how teams are using Atlan to connect every dataset, dashboard, and pipeline in our upcoming Atlan in Action session — Automated Lineage for Trust & Visibility. → Learn how automated, column-level lineage helps you predict downstream impact before changes ship → See real-time pipeline and data-quality alerts in action → Discover how OpenLineage, APIs, and custom packages extend lineage across every system 🎥 Join us live for a 45-minute interactive demo + Q&A. Save your spot 👇

Just as the data warehouse defined BI in the 90s, the context layer will define AI in the 2020s. Here's the problem: AI doesn't fail because models are wrong. It fails because AI doesn't know what humans intuitively know- the unwritten rules, edge cases, and "except when..." judgment calls that make organizations actually work. We call this the AI context gap. Recent Anthropic research confirms it: enterprises that can't effectively gather and operationalize contextual data will struggle to deploy sophisticated AI systems. But Chroma's Context Rot report also shows that more context isn't always better- beyond a certain point, it actually worsens model reasoning. The solution? Treating context as infrastructure, not an afterthought. At Atlan, we're building toward an Enterprise Context Layer- a shared foundation that captures how your organization thinks, decides, and acts. It's built on four core components: → Context extraction from existing systems → Context products (minimum viable context units) → Human-in-the-loop feedback loops → A unified context store with natural language interfaces Read the full article on how we're thinking about this: https://lnkd.in/dSaZzFts

Thrilled to share that Atlan is a launch partner for the Databricks MCP Marketplace! 🎉 This MCP Marketplace marks a major step forward in how organizations connect context, governance, and intelligence across the modern AI stack. As the metadata intelligence layer in this ecosystem, Atlan helps bring real-time context, lineage, and trust to the data powering AI agents and applications built on Databricks. We're excited to join incredible launch partners like Moody's Corporation, S&P Global, Nasdaq, GitHub, LSEG, Dun & Bradstreet, Dataiku, Bright Data, and others- all uniting around the joint vision for open, collaborative AI governance. More details in the piece linked in the comments. 🔗

Introducing the Modern Data & AI Governance Blueprint- a practitioner-first guide built from analyzing 200+ successful data governance programs. Yesterday at #ReGovern2025, we launched our first book + 32 open-source tools. And we're making it free for the entire community. Inside, you'll find: → Readiness & maturity assessments to benchmark where you stand → Step-by-step frameworks for AI-ready governance → Implementation playbooks (metadata flywheels, data product scoring) → ROI toolkits to connect governance to measurable business outcomes This is your shortcut to getting AI-ready now- not six months from now. Download here 👉 https://lnkd.in/e3XAwEz8

WHAT 👏 AN 👏 EVENT 👏 Re:Govern 2025 brought together the world's most AI-forward data teams- and one theme emerged across every conversation: context is king 👑 Some key takeaways: 1️⃣ 𝐓𝐡𝐞𝐫𝐞'𝐬 𝐧𝐨 𝐨𝐧𝐞 𝐩𝐥𝐚𝐲𝐛𝐨𝐨𝐤. CME Group and Deutsche Börse took opposite paths to Data Governance success- both worked because they matched their DNA. Meanwhile, Dropbox, Elastic, GitLab, Loopback Analytics, and Vimeo are each doing it differently, proving there's no single right answer. 2️⃣ 𝐀𝐈 𝐫𝐞𝐚𝐝𝐢𝐧𝐞𝐬𝐬 𝐬𝐭𝐚𝐫𝐭𝐬 𝐰𝐢𝐭𝐡 𝐜𝐨𝐧𝐭𝐞𝐱𝐭 𝐫𝐞𝐚𝐝𝐢𝐧𝐞𝐬𝐬. As DigiKey's CDAO put it: "Stop waiting for the business case. Build your context foundation before the crisis hits." 3️⃣ 𝐂𝐨𝐧𝐭𝐞𝐱𝐭 𝐜𝐚𝐧'𝐭 𝐛𝐞 𝐛𝐨𝐥𝐭𝐞𝐝 𝐨𝐧. From Mastercard's "context by design" to CME's automated governance at market speed, the top teams embed context from day one. 4️⃣ 𝐓𝐫𝐮𝐬𝐭 𝐞𝐧𝐚𝐛𝐥𝐞𝐬 𝐬𝐜𝐚𝐥𝐞. Virgin Media O2 turned self-service from chaos into confidence by pairing democratization with governance- reaching 16,000 people. easyJet, Nasdaq, and Invitation Homes proved that AI ROI starts with governance ROI 5️⃣ 𝐀𝐈 𝐧𝐞𝐞𝐝𝐬 𝐭𝐨 𝐛𝐞 𝐭𝐚𝐮𝐠𝐡𝐭 𝐭𝐡𝐞 𝐥𝐚𝐧𝐠𝐮𝐚𝐠𝐞 𝐨𝐟 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬. Workday, Mastercard, and CME Group showed how semantic layers turn governance into machine-readable meaning- so AI can reason, explain, and perform reliably. To every leader who shared their story: thank you for the honesty, the insight, and the inspiration. 💙 Missed Re:Govern live? Catch all the recordings here (https://lnkd.in/g5jTKjVc) or read the full recap → https://lnkd.in/g2CWN7Qp

Amie Bright from GitLab just explained why conversational analytics only works with context: "AI without context doesn't work. It's just another automation. When you add context to it, it becomes valuable." GitLab's approach: 1. Build single sources of truth (business layers) 2. Add context on top (definitions, lineage, trust signals) 3. Then unlock conversational AI The insight? "If you've invested in context and you have a platform that allows you to centralize it, then you have the ingredients for success." Conversational AI isn't magic. It's context + infrastructure.

🚨LIVE at Re:Govern 2025! Leaders from GitLab, Elastic, Dropbox, Vimeo, Loopback Analytics, easyJet, Nasdaq, Invitation Homes, New York Life Insurance Company, Mercury Insurance, and Group 1001 are sharing what it actually takes to run AI-first data organizations. Three conversations happening simultaneously: → 𝐀𝐈-𝐑𝐞𝐚𝐝𝐲 𝐓𝐞𝐜𝐡 𝐏𝐢𝐨𝐧𝐞𝐞𝐫𝐬: How fast-moving companies are preparing for AI agents, conversational analytics, and governance in decentralized cultures. → 𝐑𝐎𝐈 𝐁𝐞𝐟𝐨𝐫𝐞 𝐀𝐈: How leaders quantify the often invisible work of governance and turn foundations into strategic investments that boards understand. → 𝐄𝐯𝐨𝐥𝐮𝐭𝐢𝐨𝐧 𝐨𝐟 𝐓𝐫𝐮𝐬𝐭: How governance leaders  are moving beyond compliance and into proactive enablement using PII classification, contracts, and outcome-driven design. Join the conversation → https://lnkd.in/d8P9wTQi