Expected Databricks-Generative-AI-Engineer-Associate Exam Topics, as suggested by Databricks :

• Module 1: Design Applications: The topic focuses on designing a prompt that elicits a specifically formatted response. It also focuses on selecting model tasks to accomplish a given business requirement. Lastly, the topic covers chain components for a desired model input and output.
• Module 2: Data Preparation: Generative AI Engineers covers a chunking strategy for a given document structure and model constraints. The topic also focuses on filter extraneous content in source documents. Lastly, Generative AI Engineers also learn about extracting document content from provided source data and format.
• Module 3: Application Development: In this topic, Generative AI Engineers learn about tools needed to extract data, Langchain/similar tools, and assessing responses to identify common issues. Moreover, the topic includes questions about adjusting an LLM’s response, LLM guardrails, and the best LLM based on the attributes of the application.
• Module 4: Assembling and Deploying Applications: In this topic, Generative AI Engineers get knowledge about coding a chain using a pyfunc mode, coding a simple chain using langchain, and coding a simple chain according to requirements. Additionally, the topic focuses on basic elements needed to create a RAG application. Lastly, the topic addresses sub-topics about registering the model to Unity Catalog using MLflow.
• Module 5: Governance: Generative AI Engineers who take the exam get knowledge about masking techniques, guardrail techniques, and legal/licensing requirements in this topic.
• Module 6: Evaluation and Monitoring: This topic is all about selecting an LLM choice and key metrics. Moreover, Generative AI Engineers learn about evaluating model performance. Lastly, the topic includes sub-topics about inference logging and usage of Databricks features.

Expected Databricks-Certified-Professional-Data-Scientist Exam Topics, as suggested by Databricks :

• Module 1: A intermediate understanding of the steps in the machine learning lifecycle/ Model training, selection, and production A complete understanding of the basics of machine learning/ in-sample vs. out-of sample data
• Module 2: Applied statistics concepts/ bias-variance tradeoff
• Module 3: A complete understanding of basic machine learning algorithms and techniques/ Unsupervised techniniques like K-means and PCA
• Module 4: A complete understanding of the basics of machine learning model management/ Linear, logistic, and regularized regression
• Module 5: Tree-based models like decision trees, random forest and gradient boosted trees/ Categories of machine learning
• Module 6: Specific algorithms like ALS for recommendation and isolation forests for outlier detection/ Logging and model organization with MLflow

Expected Databricks-Certified-Professional-Data-Engineer Exam Topics, as suggested by Databricks :

• Module 1: Databricks Tooling: The Databricks Tooling topic encompasses the various features and functionalities of Delta Lake. This includes understanding the transaction log, Optimistic Concurrency Control, Delta clone, indexing optimizations, and strategies for partitioning data for optimal performance in the Databricks SQL service.
• Module 2: Data Processing: The topic covers understanding partition hints, partitioning data effectively, controlling part-file sizes, updating records, leveraging Structured Streaming and Delta Lake, implementing stream-static joins and deduplication. Additionally, it delves into utilizing Change Data Capture, and addressing performance issues related to small files.
• Module 3: Data Modeling: It focuses on understanding the objectives of data transformations, using Change Data Feed, applying Delta Lake cloning, designing multiplex bronze tables. Lastly it discusses implementing incremental processing and data quality enforcement, implementing lookup tables, and implementing Slowly Changing Dimension tables, and implementing SCD Type 0, 1, and 2 tables.
• Module 4: Security & Governance: It discusses creating Dynamic views to accomplishing data masking and using dynamic views to control access to rows and columns.
• Module 5: Monitoring & Logging: This topic includes understanding the Spark UI, inspecting event timelines and metrics, drawing conclusions from various UIs, designing systems to control cost and latency SLAs for production streaming jobs, and deploying and monitoring both streaming and batch jobs.
• Module 6: Testing & Deployment: It discusses adapting notebook dependencies to use Python file dependencies, leveraging Wheels for imports, repairing and rerunning failed jobs, creating jobs based on common use cases, designing systems to control cost and latency SLAs, configuring the Databricks CLI, and using the REST API to clone a job, trigger a run, and export the run output.

Expected Databricks-Certified-Data-Analyst-Associate Exam Topics, as suggested by Databricks :

• Module 1: Databricks SQL: This topic discusses key and side audiences, users, Databricks SQL benefits, complementing a basic Databricks SQL query, schema browser, Databricks SQL dashboards, and the purpose of Databricks SQL endpoints/warehouses. Furthermore, the delves into Serverless Databricks SQL endpoint/warehouses, trade-off between cluster size and cost for Databricks SQL endpoints/warehouses, and Partner Connect. Lastly it discusses small-file upload, connecting Databricks SQL to visualization tools, the medallion architecture, the gold layer, and the benefits of working with streaming data.
• Module 2: Data Management: The topic describes Delta Lake as a tool for managing data files, Delta Lake manages table metadata, benefits of Delta Lake within the Lakehouse, tables on Databricks, a table owner’s responsibilities, and the persistence of data. It also identifies management of a table, usage of Data Explorer by a table owner, and organization-specific considerations of PII data. Lastly, the topic it explains how the LOCATION keyword changes, usage of Data Explorer to secure data.
• Module 3: SQL in the Lakehouse: It identifies a query that retrieves data from the database, the output of a SELECT query, a benefit of having ANSI SQL, access, and clean silver-level data. It also compares and contrast MERGE INTO, INSERT TABLE, and COPY INTO. Lastly, this topic focuses on creating and applying UDFs in common scaling scenarios.
• Module 4: Data Visualization and Dashboarding: Sub-topics of this topic are about of describing how notifications are sent, how to configure and troubleshoot a basic alert, how to configure a refresh schedule, the pros and cons of sharing dashboards, how query parameters change the output, and how to change the colors of all of the visualizations. It also discusses customized data visualizations, visualization formatting, Query Based Dropdown List, and the method for sharing a dashboard.
• Module 5: Analytics applications: It describes key moments of statistical distributions, data enhancement, and the blending of data between two source applications. Moroever, the topic also explains last-mile ETL, a scenario in which data blending would be beneficial, key statistical measures, descriptive statistics, and discrete and continuous statistics.

• Module 1: Navigate the Spark UI and describe how the catalyst optimizer, partitioning, and caching affect Spark’s execution performance
• Module 2: Apply the Structured Streaming API to perform analytics on streaming data/ Define the major components of Spark architecture and execution hierarchy
• Module 3: Describe how DataFrames are built, transformed, and evaluated in Spark/ Apply the DataFrame API to explore, preprocess, join, and ingest data in Spark