_peregrine_

Karma: 166
Created: 4 years ago

About

technical writer + editor. I generally share stuff written by my peers. Currently @ turbopuffer

Recent Submissions

1. ▲ A graph-theoretic approach to building reliable LLM judges for retrieval (georgianailab.substack.com) 3 points · 22 days ago · 0 comments
2. ▲ Mixing numeric attributes into text search for better first-stage relevance (turbopuffer.com) 3 points · 1 month ago · 0 comments
3. ▲ How to build a distributed queue in a single JSON file on object storage (turbopuffer.com) 2 points · 4 months ago · 0 comments
4. ▲ ANN v3: 200ms p99 query latency over 100B vectors (turbopuffer.com) 109 points · 5 months ago · 47 comments
5. ▲ Designing inverted indexes in a KV-store on object storage (turbopuffer.com) 3 points · 5 months ago · 0 comments
6. ▲ Why BM25 queries with more terms can be faster (and other scaling surprises) (turbopuffer.com) 53 points · 5 months ago · 1 comment
7. ▲ Vectorized MAXSCORE over WAND, especially for long LLM-generated queries (turbopuffer.com) 2 points · 6 months ago · 0 comments
8. ▲ Flink's 95% Problem (tinybird.co) 2 points · 8 months ago · 0 comments
9. ▲ How to ingest 1B rows/s in ClickHouse (tinybird.co) 32 points · 10 months ago · 10 comments
10. ▲ How we built our own Claude Code (for data) (tinybird.co) 3 points · 10 months ago · 0 comments

All submissions on HN · View profile on HN