Posts

Introduction Picture this: Your marketing team needs customer data for an upcoming campaign. You submit a request to IT, initiating a complex process that involves multiple teams, approval chains, and coordination across departments. The request joins a backlog of similar requests, each requiring data team resources. This scenario plays out daily in enterprises worldwide. What should be simple data requests turn into lengthy processes that can stretch for weeks or months. ...

Introduction The data warehouse landscape is experiencing a tectonic shift. After decades of dominance by traditional vendor-locked solutions, a new way of thinking is emerging: data warehouses built on open table formats. This architectural approach isn’t just another incremental improvement—it represents a fundamental reimagining of how organizations store, manage, and analyze their critical data assets. Open table formats like Apache Iceberg, Delta Lake, and Apache Hudi are transforming what’s possible in data warehousing. By decoupling storage from compute and leveraging cloud-native technologies, these formats enable data architectures that are more flexible, cost-effective, and powerful than their traditional counterparts. ...

Introduction As data engineers and analysts, we rely heavily on SQL databases to store and query our data efficiently. To speed up our queries, we often create indexes on frequently filtered columns. However, there’s a common gotcha that can cause our queries to run slower than expected, even with appropriate indexes in place. In this post, we’ll explore how applying functions to indexed columns in the WHERE clause can prevent SQL optimizers from utilizing those indexes effectively. ...

The Problem: Union function isn’t as Simple as it Seems Picture this: You have a large dataset that you need to process in different ways, so you: Split it into smaller pieces Transform each piece differently Put them back together using union Sounds straightforward, right? Well, there’s a catch that most developers don’t know about. The Hidden Performance Killer 🐌 Here’s what’s actually happening behind the scenes when you use union: ...

Introduction When building event-driven architectures with AWS Lambda and Apache Kafka, one of the most critical yet often misunderstood components is offset management especially for event source mapping when you use lambda functions. Many developers wonder: Do I need to manage Kafka offsets manually? or What happens when my consumer group’s offsets expire? In this blog post, we’ll demystify how AWS Lambda’s Event Source Mapping handles Kafka offsets automatically and what you actually need to know as a developer. ...

DuckDB Inside Postgres: The Unlikely Duo Supercharging Analytics If you work in data engineering, you know that the field moves at a dizzying pace. New tools and technologies seem to pop up daily, each promising to revolutionize how we store, process, and analyze data. Amidst this constant change, two names have remained stalwarts: Postgres, the tried-and-true relational database, and DuckDB, the talented new kid on the block for analytics workloads. ...

Introduction In this blog post, we’ll zoom into the details of how column ordering and data types can cause issues when using the union function in Apache Spark to combine two dataframes. We’ll explore real-world examples that illustrate the problem and provide practical solutions to overcome these challenges. By the end of this post, you’ll have a better understanding of how to use union effectively and avoid common pitfalls that can lead to job failures. ...

Introduction Hi there 👋! In this blog post, we will explore why JSON is not suitable as a big data file format. We’ll compare it to the widely used Parquet format and dig deep to demonstrate, through examples, how the JSON format can significantly degrade the performance of your data processing jobs. JSON (JavaScript Object Notation) is a popular and versatile data format, but it has limitations when dealing with large-scale data operations. On the other hand, Parquet, an open-source columnar storage format, has become the go-to choice for big data applications. ...

Introduction Welcome, readers! 📖 In this post, we’ll explore the idea of event source mapping in AWS, with a focus on its implementation and functionality. We’ll zoom-in 🔍 how automatic scaling works and examine the process of consuming messages from Kafka event sources. Using Lambda to consume records from Kafka Processing streaming data with traditional server-based technologies and Kafka consumers written in Scala can often introduce unnecessary overhead for simple tasks like creating custom sink consumers to save or delete data based on specific rules. In such scenarios, where complex data processing, joining multiple topics, or a full-fledged streaming application isn’t required, a Python-based Lambda function emerges as a viable alternative. ...

Introduction Hi, welcome to my first blog post. This post is the first one in a series of many that will follow. Who is Tortoise and who is Hare? Well, In many books about Apache Spark that I was reading, I didn’t found a clear idea of the performance of dataframes compared to the datasets. In this blog post, we will debunk that mystery and show some concrete results and insights regarding this matter. ...