DSDSD - Dutch Seminar on Data Systems Design Courses

Efficient CSV Parsing - On the Complexity of Simple Things

Dive into advanced CSV parsing techniques, exploring DuckDB's implementation, parallel algorithms, buffer management, and state machine transitions for handling complex CSV file processing.

• YouTube
• 51 minutes
• On-Demand
• Free Video

Lambda Functions and List Operations in DuckDB SQL

Explore DuckDB's innovative approach to handling LIST-type operations and Python-style list comprehension in SQL, focusing on efficient processing of nested data formats like JSON for enhanced analytics performance.

• YouTube
• 25 minutes
• On-Demand
• Free Video

C3: Compressing Correlated Columns - A Framework for Multi-Column Data Compression

Discover how C3 framework revolutionizes data compression by exploiting correlations between columns, achieving 20% better compression ratios through innovative multi-column compression schemes.

• YouTube
• 25 minutes
• On-Demand
• Free Video

Towards LLM-augmented Database Systems

Explore how database systems can be enhanced with LLMs to efficiently handle multi-model data queries while addressing performance challenges and hallucination issues in modern data processing.

• YouTube
• 50 minutes
• On-Demand
• Free Video

ALP: Adaptive Lossless Floating-Point Compression

Discover advanced floating-point data compression techniques through ALP, exploring vectorized execution methods and enhanced PseudoDecimals for improved speed and compression ratios in data science applications.

• YouTube
• 29 minutes
• On-Demand
• Free Video

LingoDB: An Open Compilation and Optimization Framework for Sustainable Data Processing

Discover how LingoDB's open-source framework enables efficient data processing through flexible architecture, cross-domain optimizations, and seamless adaptation to modern hardware capabilities.

• YouTube
• 27 minutes
• On-Demand
• Free Video

Decoupling Compute and Storage for Stream Processing Systems

Explore the challenges and solutions of decoupled compute and storage in stream processing systems, focusing on tiered storage mechanisms to balance performance and cost-efficiency.

• YouTube
• 50 minutes
• On-Demand
• Free Video

Shredding Deeply Nested JSON for Efficient Analytics in DuckDB

Discover how DuckDB efficiently processes nested JSON data through vectorized analytics, enabling direct reading of JSON files as tables with automatic schema detection for powerful data analysis.

• YouTube
• 21 minutes
• On-Demand
• Free Video

Accurate Summary-based Cardinality Estimation Through the Lens of Cardinality Estimation Graphs

Explore advanced cardinality estimation techniques for graph databases, focusing on optimistic and pessimistic estimators through cardinality estimation graphs and their practical applications in query optimization.

• YouTube
• 54 minutes
• On-Demand
• Free Video

Leveraging Generative AI for Data Processing and Database Management

Explore how generative AI advances transform data management, from automated database tuning to natural language pattern mining and SQL code synthesis, with insights from Cornell's cutting-edge research.

• YouTube
• 42 minutes
• On-Demand
• Free Video

Parameter-Efficient Automation of Data Wrangling Tasks with Prefix-Tuning

Explore how prefix-tuning offers an efficient alternative to LLM fine-tuning for data wrangling tasks, requiring minimal parameter updates while maintaining comparable performance in data integration and cleaning.

• YouTube
• 17 minutes
• On-Demand
• Free Video

Implementing InfluxDB IOx Using Apache Arrow, DataFusion, and Rust

Discover how modern database systems are built using Apache Arrow, DataFusion, and Rust, exploring the architecture and implementation of InfluxDB IOx through deconstructed database approaches and open-source components.

• YouTube
• 49 minutes
• On-Demand
• Free Video

Provenance Research in Data Management Systems at Microsoft

Explore Microsoft's innovative research on provenance systems, including OneProvenance and DSProvenance engines, focusing on data lineage tracking, metadata management, and enterprise-scale data governance solutions.

• YouTube
• 44 minutes
• On-Demand
• Free Video

Database Schemas in the Wild - Learning from a Large Corpus of Relational Database Schemas

Explore the insights from GitSchemas, a vast collection of 150k+ database schemas extracted from SQL scripts, revealing patterns in real-world database design and constraints across 1M+ tables.

• YouTube
• 21 minutes
• On-Demand
• Free Video

Efficient Detection of Multivariate Correlations in Static and Streaming Data

Discover efficient algorithms for detecting multivariate correlations in data systems, exploring novel approaches that outperform existing methods by analyzing static and streaming datasets.

• YouTube
• 20 minutes
• On-Demand
• Free Video