Agentic Data Preparation for Databricks

Video Transcript: Prophecy in Action

Hi, we are the co-founders of Prophecy and we have an exciting new product to share with you. Over the past 2 years, coding without AI now seems impossible. What used to take days to weeks takes minutes. But what about data? How do we [music] bring this power of AI to millions of data users who do data prep and analysis every day? >> Let's face the reality. Most business users are not programmers. They start with spreadsheets and that works [music] until the data gets complicated and then they get stuck behind the ticketing cues of central data platforms. The backlogs pile up, the users get frustrated and business decisions get delayed. The visual desktop tools did help but they left most [music] of the companies with a wild west. Fragile logic, silos and a lot of technical debt. Today teams want and need AIdriven productivity, enterprise governance and cloud [music] scale all together. >> So we asked a simple question. How would you rebuild data prep and analysis [music] if you started with AI agents at the core? So what prophecy does is that you ask a business question. For that we prepare [music] the data and give you the analysis. To do this we have a unique approach [music] that does three main things. First is quality. We use cloud code specialized for data delivering amazing results. Second [music] is trust. You've got to be able to trust your answers and business data users [music] don't like code. So we take this code and lift it into the visual layer and the visual workflows [music] you can inspect, refine and validate. And finally we provide an end toend solution. Unlike the texttosql [music] products, we start with raw data, prepare the data for your business problem, do the analysis, do all of this on top of your cloud data platform such as data [music] bricks, snowflake or bigquery and run with scale and governance. And what do you [music] get for that? You get amazing results. Let's look at that. Let me show you how AI is changing data [music] prep and analysis. This is prophecy. On the left is the agent. This is our specialized version of cloud code. On the right is the data preparation and analysis [music] tabs. The result flow from the data preparation workflow straight into the analysis. [music] Today I'll put myself in the shoes of a marketing analyst working in a car dealership. The goal is simple. Understand historical campaigns performance and [music] launch a new campaign with better segmentation and optimized revenue targets. Let's get started. First, [music] we look at the campaign conversions. I'll point our agent to marketing data on datab bricks. Build a funnel conversion analysis for the existing [music] campaigns. Perfect. I didn't have to do a thing. AI built the whole workflow and analysis for me. The workflow reads the data from all of the stages, combines it, reshapes [music] it, and pushes it straight into the analysis, all from a single prompt. I [music] can see all of the transformations very clearly. I can validate them, and I can refine them. But there is more. [music] Here are the KPIs of the cars sold, revenue generated, and [music] we can see all of the funnel stages. This is perfect. We should update this before the daily marketing meeting. We can schedule [music] it from here. Now, let's go even further. [music] We want a new campaign with optimized segments. Let's start by creating a new workflow. Now, for the segmentation, I'm going to upload a vendor provided data set. And now let's ask the agent to recommend the good segments. uh using this demographic data recommend good segments for a marketing campaign. This is no longer just data prep. We are getting strategy and AI [music] made some great recommendations about targeting and optimizing the revenue. We can also see two clear groups affluent buyers and costconscious [music] buyers. The affluent buyers convert slightly lower, but they generate much higher revenue per sale. [music] That's because they are buying more expensive cars. The AI so far analyzed four years of data. I only want to focus on the last year for the segmentation. Let me [music] update that. And now we can see something new. California buyers are shifting toward higherend vehicles [music] than New Yorkers. Very interesting. This is [music] a gamecher. The AI not only prepped my data, but it also identified key insights and made recommendations for how to best hit [music] my business goals. And that's Prophecy. >> And how do you access Prophecy? The individuals can go to our website today and start using Prophecy for absolutely free. And for enterprises, we have Prophecy Express, [music] which will guarantee you your first production win within 90 days. >> This is big. It'll change many industries. Financial auditors will be able to go from months to near real time. [music] Marketers can build, deploy, and monitor their own campaigns. And financial asset buyers can go from raw data straight to dashboards, make the financial decisions. In 2 years, everybody will be using AI for data just as they're using AI for code. So go to prophecy.ai and get started today.