Unlock Your Day: The AI Agent That Knows What You Need

Today, our demo will focus on a fictional business, Snow Music, which manages live tour operations and high-traffic fan application. I will demo as the VP of sales, first showing how a business leader starts the day. Data everywhere, action items buried in unread emails, and hundreds of Slack messages. The information is everywhere, and another static dashboard isn't going to cut it. This is where Snowflake Co-work, your

personal work agent, comes in. Co-work brings together knowledge into one trusted enterprise agent, makes recommendations, and takes actions. Let's get started. I will begin morning

triage by prompting Co-work. Start my day. What do I need to know? Notice that there are no complex menus or agent pickers here. Exactly how a business user wants to do it. I simply ask my question, and it automatically routes the request based on my context. Snowflake Co-work is analyzing data across many source tables, emails, Slack conversations, and more. It learns my behavior over time to deliver highly personalized responses, prioritizing what actually matters. Look at this update. Tailored specifically for me. Co-work provides the critical information I should know about. While prompting this live is incredibly easy, I can also turn this into an automated flow by setting up a scheduled automation. Here, I'm showing how Co-work automations are set up. This analysis will be delivered straight into my inbox first thing in the morning, completely eliminating manual triage, allowing me to start taking actions immediately. Let's dig into something I saw in my

morning update. I noticed a surge in VIP fan upgrades at the Miami show. My regional manager has already started looking into this and shared a live interactive dashboard with me using artifacts. Artifacts now include publishable dashboards with governed views of live data. It turns one-off dashboards into reusable insights, so teams can ask questions and get insights building on each other's work. I can create this artifact directly and ask "What is driving the VIP surge in Miami?" Sophie Cork answers in line instantly with a grounded chart. Now I need to loop in our local field team. To do so, I'll simply send a Slack message to the team with this link. And just like that, the message is directly shared with via Slack using our MCP connectors to execute this task. I've diagnosed the surge, collaborated with my team, and looped in field reps without switching tools. All in one place from Cork. This VIP surge in Miami spans multiple data sources, including surveys, social sentiment, ticket resale, and venue wait times, requiring deeper exploration. To solve this, I have enabled deep

research in my prompt to execute multiple step reasoning across data and web sources in one shot. Without deep research, this would take multiple analysts and days of manual work. Because it takes 10 to 15 minutes to fully execute, I've pre-run it here. Deep research automatically spins up four parallel sub-agents to tackle different data sets at once, providing incredibly thorough research. It delivered a synthesized briefing with built-in citations. Now I understand what happened to my business and exactly why and what to do next. Next, I have a very important meeting

coming up. Usually preparing for this means chasing a bit across platforms. Instead, I'll simply ask Corg "Prepare me for my next meeting with three most important talking points." Corg securely scans meeting notes, Slack conversations, emails, and other sources to deliver a set of talking points. Now, I'm completely prepared for my meeting. To make this repeatable process, I will use the follow prompt. "Please prepare me for this meeting every Tuesday at 10:00 a.m." This automation will run every week right before my meeting. Corg is personalized for me, operating as a true productivity partner. Now, I'm going to ask a question that

spans across multiple data sets and artifacts. "Which cities had strong fan engagement but underperformed on merch in the quarter?" With Cortex Sense working under the hood, the context layer builds understanding across every artifact and data set for my organization. The more data and context the data flows through Snowflake, the richer the context becomes, and every shared asset makes the model smarter. Cortex Sense is pulling from every artifact, including fan engagement and the business artifact my GM shared earlier. Neither source alone could answer this question. Cortex Sense automatically builds that understanding for you. No manual mapping, no waiting on the data team. Look at that. It combined fan engagement data from one artifact with merch revenue from another without me telling it where to look. And because work doesn't stop when you leave your desk, this entire experience extends to a seamless mobile experience, enabling you to securely ask questions and take action on your data and workflows from anywhere on the go. To wrap up, Soul flake core is now a

personal work agent for every business user. It understands your data and your work from day one, answers the hardest why questions with cited research and runs the work on demand, on schedule, and across the tools your business already uses, all inside Soul flakes governance perimeter.