Databricks Data Science Agent: Auto-Generate & Execute Analytics

0:04Hey, my name is Yousef and today I'm going to talk about datab bricks assistant. Since it launched two years ago, it become a must-h have tool for data practitioners, helping them writing code in Python, SQL, fixing errors in workflows and much more. And now, datab bricks is raising the game by bringing agents to databicks assistant. It's a

0:26major advancement that elevates the Databricks assistant from a helpful copilot into a true autonomous partner for data science and analytics. Of course, it's fully integrated with Databricks notebooks and SQL editor. We're bringing intelligence, adaptability, and execution together in a single experience. And it can do a lot of things such as exploring data, fixing

0:48errors, summarizing, explaining results, and of course, finding relevant data. Let me show you how you can get started. Now that I am in the interface, I'm going to click on the assistant. I will write down my query. Select the table I want to analyze. It's going to be usage. And then I'll ask to extract key KPIs

1:11and build charts. Then I'll submit it to the agent. So basically the agent is asking whether I want to compute the KPIs for the last 13 days or analyze all the data. And is there any particular KPI I want to analyze? For my case, just compute the KPI 13 days. And now the s the agent will start generating the uh

1:34then the code for me and it will automatically paste it in the u in the notebook. So we're going to see popping up a bunch of them. You can also if you want accept the uh the queries. I'm going to accept them all. And now the system is going to run them. the system the permission to do so and then it will run them one after one. So the

1:57first one basically is filtering usage data for the last for the last 13 days. The other one is computing the KPI which is the total usage and the top skew top workspaces as well. And the last one is preparing data to build charts. And as I did ask when writing down my

2:19prompts, of course, I have the two results I asked. Now the prompt is going there. And as you can see, it did also add comments on the name of the cells automatically. And it did build the chart I want. Now, of course, I can click over here and I can add some dock using again the assistant that will add the uh documentation over every every

2:41single cell. I can come here again and say duck. It will write down the documentation. I can also ask to explain the uh to explain the query and it does a lot of interesting thing. It can also for optimize some workload. But you can do a lot of things. You can rename doc. You can ask for example to explain this query. It will automatically trigger the

3:03assistant and explain all the code that was written over there. So this is really the perks of using the assistant with agents. there there it just improve your productivity.