Data Engineering Podcast
Building ETL Pipelines With Generative AI
Podcast summary created with Snipd AI
Quick takeaways
- Generative AI is beneficial for handling unstructured data in ETL, providing improvements in extraction and assisting with handling text-based and image data.
- The introduction of AI in ETL is shifting responsibilities, allowing business-facing individuals to collaborate with data engineers and enabling cross-functional teams.
Deep dives
Use Cases of Generative AI in ETL
Generative AI is being applied in various areas of ETL, including data ingestion, transformation, schema mapping, and overall automation. It is particularly beneficial for handling unstructured data, such as text-based and image data, where it can assist with extraction and provide significant improvements. AI is also used to generate code for transformations, automate repetitive tasks, and enhance user experience. It is impacting decision making, and its usability is increasing with chatboard-style interfaces and natural language processing. AI plays a key role in improving efficiency, accuracy, and productivity throughout the ETL process.
Shift in User Responsibility
The introduction of AI into the ETL process is shifting the responsibilities of different roles. Business-facing individuals without technical backgrounds are now able to take on some of the tasks previously handled solely by data engineers and ETL developers. This shift allows data engineers to focus on more valuable and interesting tasks, while subject matter experts collaborate closely with them. Cross-functional teams are becoming the norm, and the dynamics of implementing data solutions are evolving.
Structured vs. Unstructured Data and AI
Generative AI is particularly impactful in dealing with unstructured data, such as text-based information, where accuracy requirements may be less strict. AI can assist in data ingestion, transformation, and extraction of valuable insights from unstructured data sources. Conversely, for structured data, AI's precision and deterministic nature may be less reliable, leading to better AI applications in automating existing ETL solutions and enhancing usability and user experience.
Considerations for AI in ETL
Before incorporating AI into ETL workflows, it is important to evaluate and strategize its application. Identifying suitable scenarios, ensuring the right tooling, and providing proper training to resources are crucial steps. Compliance and explainability concerns must be considered, choosing areas where AI can operate effectively without compromising compliance requirements. Additionally, simple use cases with structured data may not warrant the use of AI, as traditional ETL approaches can be more cost-effective and efficient.
Summary
Artificial intelligence applications require substantial high quality data, which is provided through ETL pipelines. Now that AI has reached the level of sophistication seen in the various generative models it is being used to build new ETL workflows. In this episode Jay Mishra shares his experiences and insights building ETL pipelines with the help of generative AI.
Announcements
- Hello and welcome to the Data Engineering Podcast, the show about modern data management
- Introducing RudderStack Profiles. RudderStack Profiles takes the SaaS guesswork and SQL grunt work out of building complete customer profiles so you can quickly ship actionable, enriched data to every downstream team. You specify the customer traits, then Profiles runs the joins and computations for you to create complete customer profiles. Get all of the details and try the new product today at dataengineeringpodcast.com/rudderstack
- This episode is brought to you by Datafold – a testing automation platform for data engineers that finds data quality issues before the code and data are deployed to production. Datafold leverages data-diffing to compare production and development environments and column-level lineage to show you the exact impact of every code change on data, metrics, and BI tools, keeping your team productive and stakeholders happy. Datafold integrates with dbt, the modern data stack, and seamlessly plugs in your data CI for team-wide and automated testing. If you are migrating to a modern data stack, Datafold can also help you automate data and code validation to speed up the migration. Learn more about Datafold by visiting dataengineeringpodcast.com/datafold
- You shouldn't have to throw away the database to build with fast-changing data. You should be able to keep the familiarity of SQL and the proven architecture of cloud warehouses, but swap the decades-old batch computation model for an efficient incremental engine to get complex queries that are always up-to-date. With Materialize, you can! It’s the only true SQL streaming database built from the ground up to meet the needs of modern data products. Whether it’s real-time dashboarding and analytics, personalization and segmentation or automation and alerting, Materialize gives you the ability to work with fresh, correct, and scalable results — all in a familiar SQL interface. Go to dataengineeringpodcast.com/materialize today to get 2 weeks free!
- As more people start using AI for projects, two things are clear: It’s a rapidly advancing field, but it’s tough to navigate. How can you get the best results for your use case? Instead of being subjected to a bunch of buzzword bingo, hear directly from pioneers in the developer and data science space on how they use graph tech to build AI-powered apps. . Attend the dev and ML talks at NODES 2023, a free online conference on October 26 featuring some of the brightest minds in tech. Check out the agenda and register at Neo4j.com/NODES.
- Your host is Tobias Macey and today I'm interviewing Jay Mishra about the applications for generative AI in the ETL process
Interview
- Introduction
- How did you get involved in the area of data management?
- What are the different aspects/types of ETL that you are seeing generative AI applied to?
- What kind of impact are you seeing in terms of time spent/quality of output/etc.?
- What kinds of projects are most likely to benefit from the application of generative AI?
- Can you describe what a typical workflow of using AI to build ETL workflows looks like?
- What are some of the types of errors that you are likely to experience from the AI?
- Once the pipeline is defined, what does the ongoing maintenance look like?
- Is the AI required to operate within the pipeline in perpetuity?
- For individuals/teams/organizations who are experimenting with AI in their data engineering workflows, what are the concerns/questions that they are trying to address?
- What are the most interesting, innovative, or unexpected ways that you have seen generative AI used in ETL workflows?
- What are the most interesting, unexpected, or challenging lessons that you have learned while working on ETL and generative AI?
- When is AI the wrong choice for ETL applications?
- What are your predictions for future applications of AI in ETL and other data engineering practices?
Contact Info
- @MishraJay on Twitter
Parting Question
- From your perspective, what is the biggest gap in the tooling or technology for data management today?
Closing Announcements
- Thank you for listening! Don't forget to check out our other shows. Podcast.__init__ covers the Python language, its community, and the innovative ways it is being used. The Machine Learning Podcast helps you go from idea to production with machine learning.
- Visit the site to subscribe to the show, sign up for the mailing list, and read the show notes.
- If you've learned something or tried out a project from the show then tell us about it! Email hosts@dataengineeringpodcast.com) with your story.
- To help other people find the show please leave a review on Apple Podcasts and tell your friends and co-workers
Links
- Astera
- Data Vault
- Star Schema
- OpenAI
- GPT == Generative Pre-trained Transformer
- Entity Resolution
- LLAMA
The intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA
Sponsored By:
- Materialize:  You shouldn't have to throw away the database to build with fast-changing data. Keep the familiar SQL, keep the proven architecture of cloud warehouses, but swap the decades-old batch computation model for an efficient incremental engine to get complex queries that are always up-to-date. That is Materialize, the only true SQL streaming database built from the ground up to meet the needs of modern data products: Fresh, Correct, Scalable — all in a familiar SQL UI. Built on Timely Dataflow and Differential Dataflow, open source frameworks created by cofounder Frank McSherry at Microsoft Research, Materialize is trusted by data and engineering teams at Ramp, Pluralsight, Onward and more to build real-time data products without the cost, complexity, and development time of stream processing. Go to [materialize.com](https://materialize.com/register/?utm_source=depodcast&utm_medium=paid&utm_campaign=early-access) today and get 2 weeks free!
- Datafold:  This episode is brought to you by Datafold – a testing automation platform for data engineers that finds data quality issues before the code and data are deployed to production. Datafold leverages data-diffing to compare production and development environments and column-level lineage to show you the exact impact of every code change on data, metrics, and BI tools, keeping your team productive and stakeholders happy. Datafold integrates with dbt, the modern data stack, and seamlessly plugs in your data CI for team-wide and automated testing. If you are migrating to a modern data stack, Datafold can also help you automate data and code validation to speed up the migration. Learn more about Datafold by visiting [dataengineeringpodcast.com/datafold](https://www.dataengineeringpodcast.com/datafold) today!
- Rudderstack:  Introducing RudderStack Profiles. RudderStack Profiles takes the SaaS guesswork and SQL grunt work out of building complete customer profiles so you can quickly ship actionable, enriched data to every downstream team. You specify the customer traits, then Profiles runs the joins and computations for you to create complete customer profiles. Get all of the details and try the new product today at [dataengineeringpodcast.com/rudderstack](https://www.dataengineeringpodcast.com/rudderstack)
- Neo4J:  NODES 2023 is a free online conference focused on graph-driven innovations with content for all skill levels. Its 24 hours are packed with 90 interactive technical sessions from top developers and data scientists across the world covering a broad range of topics and use cases. The event tracks: - Intelligent Applications: APIs, Libraries, and Frameworks – Tools and best practices for creating graph-powered applications and APIs with any software stack and programming language, including Java, Python, and JavaScript - Machine Learning and AI – How graph technology provides context for your data and enhances the accuracy of your AI and ML projects (e.g.: graph neural networks, responsible AI) - Visualization: Tools, Techniques, and Best Practices – Techniques and tools for exploring hidden and unknown patterns in your data and presenting complex relationships (knowledge graphs, ethical data practices, and data representation) Don’t miss your chance to hear about the latest graph-powered implementations and best practices for free on October 26 at NODES 2023. Go to [Neo4j.com/NODES](https://Neo4j.com/NODES) today to see the full agenda and register!
Remember Everything You Learn from Podcasts
