Mikkel Dengsøe’s Post

One of the key foundational elements to implementing #AI at scale is "Transforming your #data #platform from manual to automated" In my next article we discuss mapping three critical entities; #data (model, table and data elements), #asset (hosts, node and infrastructure) and #applications (ETL, ML pipelines). And creating right foundation to build and orchestrate APIs programatically. What entity type are you automating first in your data platform? I'd love to hear what's working (and what's not). #AI #Data #GenerativeAI #Transformation

Messy, incomplete user data is killing your pipeline. I just published a no‑BS guide on how to turn raw sign‑up strings into verified domains and real decision‑maker contacts, at scale, without burning engineering time on fragile scripts. I break down a production‑ready workflow using n8n + NocoDB, with Serper for domain search, GPT‑4.1‑mini for smart filtering, and Anymail finder for contact discovery. It’s built for idempotency, batching, rate limits, retries: the boring stuff that makes it bulletproof. What you’ll get: A resilient enrichment pipeline: From company name → official domain → validated contacts. Architecture + data model: Clean relational structure with Companies ↔ Contacts. Step‑by‑step n8n workflow: Ingestion, domain discovery, AI validation, contact enrichment, fallbacks, and finalization. Leader takeaways: Turn engineering from a cost center into a value driver with an internal service that feeds GTM tools directly. If your team is stuck in “quick script” hell and on‑call fires, this is the exit. Read the article: “Goodbye, messy data: An engineer’s guide to scalable data enrichment”: https://shorturl.at/l0AZ8 Special thanks to the LogRocket team! And if you want to swap ideas on automation and AI workflows, join us in the AI Alchemists community: http://ai-alchemists.com/ PS: I included the full workflow and prompts so you can clone, adapt, and ship fast.

Creating a well defined data catalog has historically been a tedious and time consuming task. With Snowflake Cortex AI this process can be streamlined, enabling efficient quality control. Check out this article demonstrating two stored procedures to be created within your Snowflake account. These procedures generate column definitions accompanied by a confidence score, ensuring accuracy before application, and ultimately establishes a self-documenting system, freeing up your team to focus on innovation rather than administrative tasks. https://lnkd.in/ent2HA9M #DataCatalog #SnowflakeCortexAI #Snowflake

If you're struggling with the "document swamp" and complex ETL, this is huge. Our team just published a solution for a fully automated, event-driven document processing pipeline—100% native to Snowflake Cortex AISQL. No Lambda functions, no external AI APIs, and no separate vector databases needed. The simplicity of AI_PARSE_DOCUMENT, AI_CLASSIFY, and AI_EXTRACT is truly a game-changer for speed and cost. This is the definition of AI without the DIY. See the full code and step-by-step implementation below.

#Couchbase 8.0, out today, delivers end-to-end AI data lifecycle support for enterprises building AI applications and agentic systems. It provides a scalable, high performing, AI-ready data platform for building context-aware, real-time AI applications. "Couchbase's new vector search capabilities transform how we deliver context-aware video discovery for enterprises. We're already using SQL++ and full-text search to query metadata across hundreds of thousands of employee-generated videos, and added vector search capabilities takes this to the next level," said Ian Merrington, CTO at Seenit. "Our customers can find relevant content based on meaning and context, not just exact keywords. As a Capella customer, we're excited for Couchbase 8.0 and the scalability and TCO benefits that make it the ideal solution for our AI-powered video platform."

If you're struggling with the "document swamp" and complex ETL, this is huge. Our team just published a solution for a fully automated, event-driven document processing pipeline—100% native to Snowflake Cortex AISQL. No Lambda functions, no external AI APIs, and no separate vector databases needed. The simplicity of AI_PARSE_DOCUMENT, AI_CLASSIFY, and AI_EXTRACT is truly a game-changer for speed and cost. This is the definition of AI without the DIY. See the full code and step-by-step implementation below. ⚡️

If you're struggling with the "document swamp" and complex ETL, this is huge. Our team just published a solution for a fully automated, event-driven document processing pipeline—100% native to Snowflake Cortex AISQL. No Lambda functions, no external AI APIs, and no separate vector databases needed. The simplicity of AI_PARSE_DOCUMENT, AI_CLASSIFY, and AI_EXTRACT is truly a game-changer for speed and cost. This is the definition of AI without the DIY. See the full code and step-by-step implementation below. ⚡️

If you're struggling with the "document swamp" and complex ETL, this is huge. Our team just published a solution for a fully automated, event-driven document processing pipeline—100% native to Snowflake Cortex AISQL. No Lambda functions, no external AI APIs, and no separate vector databases needed. The simplicity of AI_PARSE_DOCUMENT, AI_CLASSIFY, and AI_EXTRACT is truly a game-changer for speed and cost. This is the definition of AI without the DIY. See the full code and step-by-step implementation below. ⚡️

AI Overload: Yes, we all get it, AI is awesome. And everyone wants to sell you their version of AI awesome. So how do you pick the right AI product? Is it like Bitcoin, where you can't go wrong just sticking with the OG, OpenAI? (Hot take, I know... there are so many valuable alt coins out there.) When it comes to AI products, we all know that 90% won't make it. So.... Which tools are really worth your time to investigate? Easy answer use the tools built around the products you are already using. For example, Sebastian Cajamarca built an AI agent that can be attached to any open-source version of Metabase. Every tech startup using Metabase should consider his tool. As the VP of Engineering at Pathpoint, we used Metabase for all of our BI reports. It was my first introduction to the product. I had used Microsoft Power BI, Tableau, and Pentaho in the past, but I was impressed by how powerful Metabase was for a free open-source tool. That's why I chose Metabase as our BI Visualization tool when I started building our DataXcel product at Xcel Software. As I was building our own AI Agent to attach to the Metabase Platform, it occurred to me that someone else might be doing the same thing, and be farther along, and better at it than I was. I was intrigued and excited to find Sebastian Cajamarca's work and started following his progress on Medium. About 2 months ago, I reached out to Sebastian Cajamarca about forming a partnership to help each other grow our companies. If you are using Metabase and want to leverage the data, reports, and infrastructure you've already created, and put its AI agent to work, you will be surprised at how easy it is to set up and get instant value. I've attached his latest Medium article; it's a must-read if you're using open source Metabase. Especially if you're considering moving to the enterprise version of Metabase to get the AI features they recently released. If you need help setting up Metabase, MetAgent, or a data warehouse, I'm happy to assist you.

"Many data engineers are using one tool to collect data, one tool to process data and another to run analytics on that data. Using several tools along this data lifecycle introduces complexity, risk and increased infrastructure management, which data engineers can't afford to take on." — Christopher Child, VP of product for data engineering, Snowflake 🗣️ “The result is a productivity paradox. AI tools are making individual tasks faster, but the proliferation of disconnected tools is making the overall system more complex to manage. For enterprises racing to deploy AI at scale, this fragmentation represents a critical bottleneck.” by Sean M. Kerner, VentureBeat ✍️👨💻📰🗞️ https://lnkd.in/eKKu8fY4