# Performance Articles

> Performance engineering for analytics: Query optimization, columnar storage, and vectorized execution. Real benchmarks and production latency wins.

Filter by Topic:

[All Posts (45)](/blog.md) [Engineering (31)](/blog/tag/engineering.md) [Architecture (21)](/blog/tag/architecture.md) [Performance (14)](/blog/tag/performance.md) [Product (13)](/blog/tag/product.md) [Altertable Agent (12)](/blog/tag/ai-agents.md) [Open Source (10)](/blog/tag/open-source.md) [Data Stack (9)](/blog/tag/data-stack.md) [Analytics (7)](/blog/tag/analytics.md) [AI (4)](/blog/tag/ai.md) [Culture (3)](/blog/tag/culture.md) [Generative UI (2)](/blog/tag/generative-ui.md) [Agents (2)](/blog/tag/agents.md) [Semantic Layer (2)](/blog/tag/semantic-layer.md) [Cousteau (1)](/blog/tag/cousteau.md) [MCP Apps (1)](/blog/tag/mcp-apps.md) [Agentic UI (1)](/blog/tag/agentic-ui.md) [Model Context Protocol (1)](/blog/tag/model-context-protocol.md) [Pricing (1)](/blog/tag/pricing.md)



Performance engineering for analytics: Query optimization, columnar storage, and vectorized execution. Real benchmarks and production latency wins.

[![The Best Time to Go Deep](/assets/blog/2026-10-27-the-best-time-to-go-deep.jpg)](/blog/2026-10-27-the-best-time-to-go-deep.md)

[OCTOBER 27TH, 2026](/blog/2026-10-27-the-best-time-to-go-deep.md)

[Sylvain Utard](/blog/2026-10-27-the-best-time-to-go-deep.md)

## [The Best Time to Go Deep](/blog/2026-10-27-the-best-time-to-go-deep.md)

[Engineering, Performance, Open Source, Altertable Agent](/blog/2026-10-27-the-best-time-to-go-deep.md)

[Agents made systems iterations cheap: in three days we ported a concurrent B-tree to C++ for DuckDB. Expertise still decided what to chase.](/blog/2026-10-27-the-best-time-to-go-deep.md)

[READ ARTICLE](/blog/2026-10-27-the-best-time-to-go-deep.md)

[![Event Paths in DuckDB](/assets/blog/2026-09-17-journeys-in-duckdb.jpg)](/blog/2026-09-17-journeys-in-duckdb.md)

[SEPTEMBER 17TH, 2026](/blog/2026-09-17-journeys-in-duckdb.md)

[Sylvain Utard](/blog/2026-09-17-journeys-in-duckdb.md)

## [Event Paths in DuckDB](/blog/2026-09-17-journeys-in-duckdb.md)

[Performance, Engineering, Open Source, Analytics](/blog/2026-09-17-journeys-in-duckdb.md)

[Discover observed paths through event data with JOURNEYS, a native DuckDB operator for start-to-end routes including incomplete journeys.](/blog/2026-09-17-journeys-in-duckdb.md)

[READ ARTICLE](/blog/2026-09-17-journeys-in-duckdb.md)

[![Session Windows in DuckDB](/assets/blog/2026-09-01-session-windows-in-duckdb.jpg)](/blog/2026-09-01-session-windows-in-duckdb.md)

[SEPTEMBER 1ST, 2026](/blog/2026-09-01-session-windows-in-duckdb.md)

[Sylvain Utard](/blog/2026-09-01-session-windows-in-duckdb.md)

## [Session Windows in DuckDB](/blog/2026-09-01-session-windows-in-duckdb.md)

[Performance, Engineering, Open Source](/blog/2026-09-01-session-windows-in-duckdb.md)

[Session windows group events by inactivity gaps. DuckDB's SESSIONIZE clause runs 3.4× faster than equivalent window functions on 10 million rows.](/blog/2026-09-01-session-windows-in-duckdb.md)

[READ ARTICLE](/blog/2026-09-01-session-windows-in-duckdb.md)

[![MATCH\_RECOGNIZE That Scales](/assets/blog/2026-08-07-match-recognize-that-scales.jpg)](/blog/2026-08-07-match-recognize-that-scales.md)

[AUGUST 7TH, 2026](/blog/2026-08-07-match-recognize-that-scales.md)

[Sylvain Utard](/blog/2026-08-07-match-recognize-that-scales.md)

## [MATCH\_RECOGNIZE That Scales](/blog/2026-08-07-match-recognize-that-scales.md)

[Performance, Engineering, Open Source](/blog/2026-08-07-match-recognize-that-scales.md)

[Why funnels belong in MATCH\_RECOGNIZE, how active-run automata get expensive, and how our DuckDB extension stays fast against Trino.](/blog/2026-08-07-match-recognize-that-scales.md)

[READ ARTICLE](/blog/2026-08-07-match-recognize-that-scales.md)

[![Know What a Query Reads Before It Runs](/assets/blog/2026-06-15-know-what-a-query-reads.jpg)](/blog/2026-06-15-know-what-a-query-reads.md)

[JUNE 15TH, 2026](/blog/2026-06-15-know-what-a-query-reads.md)

[Florian Valeye](/blog/2026-06-15-know-what-a-query-reads.md)

## [Know What a Query Reads Before It Runs](/blog/2026-06-15-know-what-a-query-reads.md)

[Architecture, Altertable Agent, Performance](/blog/2026-06-15-know-what-a-query-reads.md)

[Row estimates only hint at what an agent will read. Bytes pin it down. Our explain returns them from catalog metadata, before the engine touches a file.](/blog/2026-06-15-know-what-a-query-reads.md)

[READ ARTICLE](/blog/2026-06-15-know-what-a-query-reads.md)

[![Grep your lakehouse](/assets/blog/2026-03-27-grep-your-lakehouse.jpg)](/blog/2026-03-27-grep-your-lakehouse.md)

[MARCH 27TH, 2026](/blog/2026-03-27-grep-your-lakehouse.md)

[Sylvain Utard](/blog/2026-03-27-grep-your-lakehouse.md)

## [Grep your lakehouse](/blog/2026-03-27-grep-your-lakehouse.md)

[Product, Performance, Engineering](/blog/2026-03-27-grep-your-lakehouse.md)

[Agents fail when they cannot retrieve the right data slice before writing SQL—not because they cannot generate queries.](/blog/2026-03-27-grep-your-lakehouse.md)

[READ ARTICLE](/blog/2026-03-27-grep-your-lakehouse.md)

[![AI's Event Backbone](/assets/blog/2026-03-10-ai-s-events-backbone.jpg)](/blog/2026-03-10-ai-s-events-backbone.md)

[MARCH 10TH, 2026](/blog/2026-03-10-ai-s-events-backbone.md)

[Sylvain Utard](/blog/2026-03-10-ai-s-events-backbone.md)

## [AI's Event Backbone](/blog/2026-03-10-ai-s-events-backbone.md)

[Product, Performance, Engineering](/blog/2026-03-10-ai-s-events-backbone.md)

[AI-native products generate a new kind of infrastructure problem. Here's how to build the event backbone for your AI system.](/blog/2026-03-10-ai-s-events-backbone.md)

[READ ARTICLE](/blog/2026-03-10-ai-s-events-backbone.md)

[![One Billion Rows](/assets/blog/2026-02-17-one-billion-rows.jpg)](/blog/2026-02-17-one-billion-rows.md)

[FEBRUARY 17TH, 2026](/blog/2026-02-17-one-billion-rows.md)

[Sylvain Utard](/blog/2026-02-17-one-billion-rows.md)

## [One Billion Rows](/blog/2026-02-17-one-billion-rows.md)

[Product, Performance, Engineering](/blog/2026-02-17-one-billion-rows.md)

[At 1 billion rows, every shortcut comes back to collect interest. Here's how we achieved sub-second queries with near-realtime ingestion.](/blog/2026-02-17-one-billion-rows.md)

[READ ARTICLE](/blog/2026-02-17-one-billion-rows.md)

[![Pruning Top-N Queries](/assets/blog/2026-02-03-pruning-top-n-queries.jpg)](/blog/2026-02-03-pruning-top-n-queries.md)

[FEBRUARY 3RD, 2026](/blog/2026-02-03-pruning-top-n-queries.md)

[Sylvain Utard](/blog/2026-02-03-pruning-top-n-queries.md)

## [Pruning Top-N Queries](/blog/2026-02-03-pruning-top-n-queries.md)

[Open Source, Performance, Architecture](/blog/2026-02-03-pruning-top-n-queries.md)

[A deep dive into DuckLake PR #668 and how Top-N dynamic filter pruning turns ORDER BY + LIMIT from full scans into metadata-driven execution.](/blog/2026-02-03-pruning-top-n-queries.md)

[READ ARTICLE](/blog/2026-02-03-pruning-top-n-queries.md)

[![Lessons from Search](/assets/blog/2026-01-13-lessons-from-search.jpg)](/blog/2026-01-13-lessons-from-search.md)

[JANUARY 13TH, 2026](/blog/2026-01-13-lessons-from-search.md)

[Sylvain Utard](/blog/2026-01-13-lessons-from-search.md)

## [Lessons from Search](/blog/2026-01-13-lessons-from-search.md)

[Performance, Architecture, Engineering](/blog/2026-01-13-lessons-from-search.md)

[Real-time analytics faces the small-file problem search engines solved. DuckLake's tiered compaction brings those merge strategies to streaming analytics.](/blog/2026-01-13-lessons-from-search.md)

[READ ARTICLE](/blog/2026-01-13-lessons-from-search.md)

[![Stop Batching Analytics](/assets/blog/2025-12-30-stop-batching-analytics.jpg)](/blog/2025-12-30-stop-batching-analytics.md)

[DECEMBER 30TH, 2025](/blog/2025-12-30-stop-batching-analytics.md)

[Sylvain Utard](/blog/2025-12-30-stop-batching-analytics.md)

## [Stop Batching Analytics](/blog/2025-12-30-stop-batching-analytics.md)

[Analytics, Architecture, Performance](/blog/2025-12-30-stop-batching-analytics.md)

[Why we're forcing analytics through complex batch pipelines when append-only data should work like logs. The warehouse constraint that stopped making sense.](/blog/2025-12-30-stop-batching-analytics.md)

[READ ARTICLE](/blog/2025-12-30-stop-batching-analytics.md)

[![Speed Shapes Understanding](/assets/blog/2025-10-07-speed-shapes-understanding.jpg)](/blog/2025-10-07-speed-shapes-understanding.md)

[OCTOBER 7TH, 2025](/blog/2025-10-07-speed-shapes-understanding.md)

[Sylvain Utard](/blog/2025-10-07-speed-shapes-understanding.md)

## [Speed Shapes Understanding](/blog/2025-10-07-speed-shapes-understanding.md)

[Performance, Product](/blog/2025-10-07-speed-shapes-understanding.md)

[Speed isn't just a luxury: it's the difference between insight and inertia. We've been deep in TPC-H benchmarks, tuning our analytical engine for AI agents.](/blog/2025-10-07-speed-shapes-understanding.md)

[READ ARTICLE](/blog/2025-10-07-speed-shapes-understanding.md)

[![NetQuack 4000x Faster](/assets/blog/2025-09-30-netquack-4000x-faster.jpg)](/blog/2025-09-30-netquack-4000x-faster.md)

[SEPTEMBER 30TH, 2025](/blog/2025-09-30-netquack-4000x-faster.md)

[Sylvain Utard](/blog/2025-09-30-netquack-4000x-faster.md)

## [NetQuack 4000x Faster](/blog/2025-09-30-netquack-4000x-faster.md)

[Performance, Engineering](/blog/2025-09-30-netquack-4000x-faster.md)

[We rewrote NetQuack DuckDB extension, replacing regex with character parsing. Result: 4000x faster—37 seconds down to 0.012 seconds.](/blog/2025-09-30-netquack-4000x-faster.md)

[READ ARTICLE](/blog/2025-09-30-netquack-4000x-faster.md)

[![Rethinking the Lakehouse](/assets/blog/2025-07-30-rethinking-the-lakehouse.jpg)](/blog/2025-07-30-rethinking-the-lakehouse.md)

[JULY 30TH, 2025](/blog/2025-07-30-rethinking-the-lakehouse.md)

[Yannick Utard](/blog/2025-07-30-rethinking-the-lakehouse.md)

## [Rethinking the Lakehouse](/blog/2025-07-30-rethinking-the-lakehouse.md)

[Architecture, Performance, Data Stack](/blog/2025-07-30-rethinking-the-lakehouse.md)

[Breaking down our storage and query architecture: why we're leaning into Apache Iceberg and why DuckDB is emerging as our real-time query engine of choice.](/blog/2025-07-30-rethinking-the-lakehouse.md)

[READ ARTICLE](/blog/2025-07-30-rethinking-the-lakehouse.md)

## Interested in learning more about Altertable?

[Start free](/sign-up)

[Web version](https://altertable.ai/blog/tag/performance)
