Lakegres: Vector search on existing Open/Hudi/Iceberg tables
Comments on “Lakegres: Vector search on existing Open/Hudi/Iceberg tables”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
Onehouse added vector search to Lakegres, storing an IVF-style index with RaBitQ quantization in the metadata of existing Hudi and Iceberg tables, so Postgres clients can query lake data without copying it or rewriting files. On a 247M-vector, 1,024-dimension table, it reports warm p50 of 63.5 ms against 655 ms for Databricks AI Search, at similar recall (about 0.95). These are the vendor's own figures, from one workload, with different latency measurement methods, and the "only open index" claim is unsourced.
Why it ranked
No. 16 of 144 with a score of 0.65. The panel rated it highest on technical depth (0.93) and new information (0.93), lowest on audience breadth (0.60). Placed No. 16, No. 33 by score alone: the placement terms moved other stories past it.
- AI judgeJudge 2claude-haiku-5-5
- Detailed technical product launch on vector search over open lakehouse tables with a benchmark and open RFC; strong technical depth, though the benchmark is vendor-run and some claims are self-comparative and promotional in tone.
- AI judgeJudge 3gpt-5.6-luna
- A technically detailed vector-index release for open lakehouse tables, with architecture, consistency behavior, SQL usage, reproducible scale benchmarks, and caveats. It is highly actionable, though vendor-produced comparisons and future scalability claims warrant some skepticism.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.65total under rubric v2.3
- Significance0.68
- Reader interest0.81
- Audience breadth0.60
- Practical relevance0.78
- Substantiated0.78
- Technical depth0.93
- New information0.93
- Counts against: Promotional0.30
- Counts against: Overclaim0.41
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.71 | 0.55 | 0.78 | 0.68 | 0.23 | +0.24 | +0.163 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.73 | 0.80 | 0.90 | 0.81 | 0.17 | +0.26 | +0.211 |
| Audience breadthHow many people who build technology does it concern? | 0.57 | 0.60 | 0.62 | 0.60 | 0.05 | +0.13 | +0.078 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.82 | 0.70 | 0.82 | 0.78 | 0.13 | +0.22 | +0.172 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.93 | 0.60 | 0.82 | 0.78 | 0.33 | +0.05 | +0.039 |
| Technical depthHow technically substantive is it? | 0.98 | 0.85 | 0.95 | 0.93 | 0.13 | +0.05 | +0.046 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.90 | 0.90 | 0.98 | 0.93 | 0.08 | +0.05 | +0.046 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.12 | 0.50 | 0.28 | 0.30 | 0.38 | −0.19 | −0.057 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.30 | 0.55 | 0.38 | 0.41 | 0.25 | −0.11 | −0.045 |
| Sum of contributions | +0.653 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.653 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-19/16/"><img src="https://rubric.hrih.si/badge/2026-11-19/16.svg" alt="Ranked #16 on Rubric, 19 Nov 2026" width="245" height="28"></a>Claim check
Claim check: 3 matches source · 4 no source given · 1 couldn’t check. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
No source givenBenchmark result
On a 247M-vector, 1,024-dimension table, Lakegres returned warm queries in 63.5 ms at p50, 67.9 ms at p90 and 75.3 ms at p99, while Databricks AI Search took 655 ms, 876 ms and 1,209 ms.
The article gives no link or citation for it.
The article says “For warm queries, Lakegres returned results in 63.5 ms at p50, 67.9 ms at p90 and 75.3 ms at p99. Databricks AI Search took 655 ms, 876 ms and 1,209…”
The source says No source to compare with.
No source givenBenchmark result
Cold p50 latency was 536.2 ms for Lakegres and 8,242 ms for Databricks AI Search; Recall@10 was 0.95 for Lakegres and 0.952 for AI Search.
The article gives no link or citation for it.
The article says “Cold p50 latency was 536.2 ms for Lakegres and 8,242 ms for AI Search. Recall@10 was 0.95 for Lakegres and 0.952 for AI Search”
The source says No source to compare with.
No source givenBenchmark result
Building the index took 18 minutes on Lakegres and about 3 hours 10 minutes on Databricks AI Search.
The article gives no link or citation for it.
The article says “building the index took 18 minutes on Lakegres and about 3 hours 10 minutes on AI Search”
The source says No source to compare with.
Couldn’t checkNumber
The AWS instances used for Lakegres cost about $5.79/hour at US East on-demand rates, excluding Onehouse fees, versus $7.68/hour for the six AI Search units.
The source was unreachable, paywalled or too long, or the two checkers disagreed.
The article says “The AWS instances cost about $5.79/hour at US East on-demand rates, excluding Onehouse fees; the six AI Search units cost $7.68/hour”
The source says No passage could be compared.
Source cited by the article: pricing.us-east-1.amazonaws.com/offers/v1.0/aws/AmazonEC2/current/us-east-1/index.csv
Matches sourceNumber
Per AWS's sizing formula for OpenSearch HNSW, 247 million 1,024-dimension vectors with m=16 need about 1.15 TB of memory, and about 46 TB at 10 billion vectors.
The cited source says the same thing.
The article says “With m set to 16, that comes to about 1.15 TB of memory for one table, and about 46 TB at 10 billion vectors”
The source says “each HNSW graph with 32-bit float data takes approximately 1.1 * (4 * d + 8 * m) * num_vectors bytes of memory”
Source cited by the article: aws.amazon.com/blogs/big-data/cost-optimized-vector-database-introduction-to-amazon-opensearch-service-quantization-techniques
No source givenFirst, only or fastest
Lakegres is the only available open vector search index for existing Hudi and Iceberg tables that is maintained with table writes and leaves data files unchanged.
The article gives no link or citation for it.
The article says “this is the only available open vector search index for existing Hudi and Iceberg tables that is maintained with table writes and leaves the data files unchanged”
The source says No source to compare with.
Matches sourceNumber
Uber's data lake holds 350 petabytes and ingests 6 trillion rows a day.
The cited source says the same thing.
The article says “Uber’s lake holds 350 petabytes and ingests 6 trillion rows a day”
The source says “6 trillion rows across ingested per day and 3 million new data files added”
Source cited by the article: uber.com/blog/apache-hudi-at-uber
Matches sourceGeneralization
Snowflake does not support VECTOR columns on Iceberg tables.
The cited source says the same thing.
The article says “VECTOR columns are not supported on Iceberg tables”
The source says “The VECTOR data type isn’t supported for use with the following Snowflake features: - Snowflake Scripting - Apache Iceberg™ tables”
Source cited by the article: docs.snowflake.com/en/sql-reference/data-types-vector
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 40 pts, front page
Recorded for context only, checked 19 Nov 2026, 05:47 UTC. It played no part in the ranking.