Streaming one zettabyte of HD video would take over 30 million years 🤯 … so who's reading all the data our machines generate? Let’s take a tech walk @splunk .conf26 with Seth Brickman (VP Global Product Adoption at @Cisco ) as he breaks it down ⚡️

Andrew Ng (@AndrewYNg ) and the @DeepLearningAI team analyzed 10,000+ job postings + expert interviews and just shared a map of the skills that matter most in AI engineering right now for devs ⬇️

⚡ Lakebase vs Lakehouse in < 90 seconds. Why do you need both? Michael Armbrust (@michaelarmbrust ), Distinguished Engineer at @databricks, is back for another tech walk to break it down

The 4 main types of Agent Memory, explained in 40 seconds by @richmondalake (Director of AI DevEx at Oracle) ⚡️

Google engineer explains Agent Evaluation vs Model Evaluation in 30 seconds ⚡

⚡ What are Declarative Pipelines in < 90 seconds? Let's walk with @michaelarmbrust , Distinguished Engineer @databricks who led the design & development of Declarative Pipelines + created Spark SQL, Structured Streaming, Delta Lake, & much more as he breaks it down! 💡 TLDR: instead of telling the system HOW to do every step, you just tell it WHAT you want to exist. Spark figures out the how. The technology itself is open source in Apache Spark, and Databricks offers a managed version on top called Lakeflow Spark Declarative Pipelines It was such a blast getting to chat with Michael! Did you know he was also the 9th employee at Databricks?!

Datadog was born out of the friction between dev and ops. Let's walk with co-founder & CTO Alexis Lê-Quôc as he breaks down the friction he sees today Love the way Alexis phrased this: "Friction never comes from ill intent. It comes from speaking a slightly different language." 💡That's the problem Alexis and Olivier set out to solve. They'd stood on opposite sides of the dev and ops divide and seen how hard it was to get two teams to understand their systems the same way. So in 2010 they built @datadoghq to break down those silos across systems, teams, and functions, giving every team one shared view of the same data. What started as infrastructure monitoring grew into observability across the whole stack, & today the same idea extends to AI, which is where a lot of DASH 2026 (which i recently attended) pointed: giving teams visibility into what their AI agents are actually doing ✨

Tech talk & walk… the dog. 🐕 Walk with @simba_khadder, Choops 🐾, & me as he explains Redis Iris in 30 seconds! Redis Iris is Redis’s new unified, real-time context engine for AI agents. It’s made up of a few components: 🔹𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗲𝗿: turns your business data into structured tools agents call directly, so they never touch your database and can only use what you’ve explicitly defined. 🔹𝗔𝗴𝗲𝗻𝘁 𝗠𝗲𝗺𝗼𝗿𝘆: gives agents short-term session memory plus long-term memory across interactions, so context compounds instead of resetting every session. 🔹𝗟𝗮𝗻𝗴𝗖𝗮𝗰𝗵𝗲: keeps repeated and semantically similar LLM work fast by serving cached responses from Redis inside the agent’s latency budget. 🔗Check out Redis Iris here: fandf.co/4eggv7C

There has been an 800% increase in FDE (forward deployed engineer) job postings in the past year 🤯 So I had to ask the INCREDIBLE Leah McGowen, SVP of Forward-Deployed Engineering at Salesforce .... whose team tripled in just six months btw .... how she actually defines the role and what they look for when they hire. Last month I attended Agentforce World Tour NY @salesforce & her keynote was absolutely fantastic!!! One of the best speakers I've seen. I was fangirling 🥹 An FDE is a deep technical expert - part tech guru, part business consultant. Leah's description: they're the unicorns on the front line with customers, with a direct line back to product so what they learn in the field shapes the roadmap. And who they hire for it? The curious ones. The constant learners and tinkerers who want to stay on the brink of the newest tech. Thank you so much to Leah for taking the time to chat with me! #SalesforcePartner

Traditional RAG vs Graph RAG explained in 60 seconds by Nacho Martinez (Data Scientist Advocate @OracleDevs ) ⚡️ 💡 Both pull external knowledge into an LLM. The difference is whether what comes back is isolated, or actually connected. Traditional RAG returns the top similarity matches. Each chunk stands alone. Graph RAG organizes knowledge as a graph with edges between related items. Retrieval can follow those edges, so you get a match plus what it connects to. Nacho Martínez walks through it at the Oracle office 🏢 ... also pretty sure the architecture looks like databases haha

Walk with @robertnishihara & I in NYC with 10% charge 🪫 as we talk through 5 key differences between 𝗟𝗟𝗠 𝗶𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗩𝗦 𝗥𝗲𝗴𝘂𝗹𝗮𝗿 𝗶𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 Let’s see how much we can get through before our mic dies! 🤣

Context vs Memory vs Harness engineering explained in 40 seconds by @richmondalake ⚡️ 💡These 3 disciplines are core to building agents that can actually remember and reason over time Most agents work well within a single session but lose everything the moment it ends. Memory engineering treats long-term memory as first-class infrastructure. Richmond Alake (Director of AI DevEx at Oracle) walked me through all 3 in NYC🗽