GOTO - The Brightest Minds in Tech Building Modern Databases with the FDAP Stack • Andrew Lamb & Olimpiu Pop
Jan 20, 2026
Andrew Lamb, a Staff Engineer at InfluxData and Apache PMC member, joins tech journalist Olimpiu Pop to explore the innovative FDAP Stack. They dive into the shift from row-based to columnar storage and discuss how technologies like Apache Iceberg improve interoperability. Andrew shares insights on Arrow's standardized memory model, the advantages of Parquet over traditional formats, and the role of open source in accelerating database development. Their conversation highlights the future of database architecture through modular components.
AI Snips
Chapters
Transcript
Episode notes
Don’t Rebuild Core Database Tech
- Rebuilding a database from scratch is a multi-year, multi-team effort that duplicates huge prior work.
- Using existing open components lets teams focus and avoid reinventing core database systems.
Columnar Storage Enables Fast Analytics
- Columnar storage optimizes analytic workloads by grouping same-typed values for fast vectorized processing.
- Apache Arrow standardizes in-memory columnar layouts so systems interoperate and avoid repeated low-level design work.
Use Arrow To Avoid Serialization Pain
- Adopt Arrow to pass columnar data between systems without serialization overhead.
- Rely on its agreed formats to avoid spending engineering effort solving low-level representation problems.