Have you ever struggled to create sample SAP tables or find meaningful data online? Navigating SAP data can be challenging as it requires a blend of technical know-how, an understanding of business processes, and familiarity with SAP systems.
DataOps.live Assist is your go-to tool! With its mix of artificial intelligence and human intuition, creating valuable SAP insights on Snowflake has never been simpler.
Join me as we explore the magic of DataOps.live Assist. See how Assist simplifies the process of generating sample SAP data without having any knowledge about SAP. At the end of this blog post, you can access a video demonstration to complement it.
Let's step into a scenario where we create sample SAP models—KNA1, KNB1, KNVI, KNVV, VBAK, and VBAP—focusing on customer and sales data. These table names may seem cryptic at first glance, but each holds valuable information within the SAP systems, such as the SAP R\3 ERP system.
We’ll then merge these tables to generate a comprehensive view of customer and sales insights. Throughout this process, we'll interact with Assist by asking questions in straightforward human language or providing prompts to achieve our desired outcome.
Understanding cryptic and obscure SAP metadata
Assist can understand SAP data, even when the table names and columns are obscure. It can interpret human questions, understand the semantic meaning behind them, and craft advanced SQL queries to extract relevant insights.
With just a simple question such as “What SAP tables do you know about?" Assist can provide you with a business-related response showing how SAP data is organized and structured in various modules within the system, such as Master Data (MD), Finance (FI), Controlling (CO), Sales and Distribution (SD), and Material Management (MM). 
Creating meaningful sample SAP data
Generating meaningful sample SAP data can be a tough task. However, with Assist this process becomes simple. Assist automatically generates example SAP tables filled with thousands of rows of sample data. It ensures that the data is referentially correct, making it suitable for various analytical tasks.
Let's consider the scenario where we task Assist with generating customer master data. Achieving this requires just a minimal prompt and a few hints applicable to any sample data generation process.
For example: “Create a dbt model to create SAP KNA1 table with all the possible columns that usually exist for a table within SAP and 1000 rows of sample data, each of the right data type, using the Snowflake SEQ4 GENERATOR and TABLE functions without WINDOW functions. Don't use ROW_NUMBER or WINDOW functions. Use SEQ4() to generate the ID. Ensure no duplicate columns. Include the prompt to create the model at the top as documentation in the dbt config block.”
Building SAP models
After creating the sample preview, Assist creates the SAP model for KNA1 with the necessary columns. Once we run the generated SQL query and the model, it creates sample data, too.
Next, we use the same approach to create:
- KNB1, ensuring it includes all the columns needed to join the previously generated KNA1 table.
- KNVI, using Assist’s increased context to replicate the process for a specific table name as before.
- KNVV, VBAK, and VBAP, where Assist even predicts our intention to create these tables. Consequently, we continue with the same process for these other tables.
As a result, we now have models that generate a thousand rows of sample data, all running inside Snowflake without the need for complex SQL queries or SAP expertise. 
Visualizing SAP tables relationships
With just a simple prompt to create an ERD in mermaid format for the models and their columns, Assist generates a diagram showing the connection between the models and their columns.
Now, we have a clear understanding of their relationship: KNB1, KNVI, and KNV are directly linked to KNA1, while VBAP is connected through VBAK.
Joining SAP tables
With the prompt “create a model that joins all 6 tables together”—or any similar straightforward request—Assist integrates these tables, initiating the creation of a unified model. It accurately interprets the request and offers a comprehensive view of customer and sales data. It shows its ability to understand the significance of these tables and infer the insights gained from joining them.
Upon previewing the model, we now have a cohesive customer and sales table, affirming the accuracy of our join keys. 
Creating tests
By simply asking Assist to help in plain English, we can generate a test file for each model, incorporating referential integrity checks. This critical step guarantees the accuracy and reliability of our dataset.
Assist generates the test files for the SAP models, enabling us to validate the data with confidence. With the assurance of the data integrity, we now have a valuable dataset to support our demos. 
Conclusion
DataOps.live Assist not only understood SAP data but also generated cohesive sample tables quickly. It simplifies the entire process with its automation capabilities for sample data generation, model construction, table joining, and test creation. With Assist, you can focus on deriving valuable insights from SAP data without the need for specialized SAP knowledge or expertise.
Ready to experience the power of DataOps.live Assist for yourself? Watch this video for a visual walkthrough of this blog post and unlock the full potential of your SAP data on Snowflake. Get started today to speed up and start enjoying the way you extract value from your data.
Discover how Assist unlocks SAP insights on Snowflake in our newest blog post, "How DataOps.live Assist Creates SAP Insights on Snowflake in Seconds.”
Schedule a demo to see the full DataOps.live platform.
