Skip to main content
pgvector SignalsBack to the read

Every source behind the read

Every source GILD analyzed, newest first. Each says who the source is, how it knows, whether it still applies and which findings it counts toward. Open a count to see only its sources.

Whose sources count

Source

31 sources in Teams and technical leaders, with product documentation Showing all 33 sources

  1. Negative

    Nikolai Maasø Lunde (nlundee), Digdir (GitHub profile)

    GitHub issue by nlundeeGitHub issue

    Azure ships pgvector 0.8.2, where VACUUM running beside inserts can corrupt an HNSW index without an error. The issue plans to turn autovacuum off on HNSW-indexed tables until Azure ships 0.8.4 or later.

    Published Oct 29, 2026self-described: Genki, Digdir's shared generative AI platform for the Norwegian public sector, on Azure

    Counts toward

    Published on
    Developer Communities
    Who
    Organization-wide
    Evidence
    Reported
    Applies to
    Current

  2. Negative

    jito (jitokim), Applied AI Engineer, Lbox (GitHub profile)

    GitHub issue by jitokimGitHub issue

    In production, per-row insert cost on a sparsevec HNSW index regressed 38x, from 1,107 to 42,403 buffers per row, starting about 29 hours after the first autovacuum. The maintainer merged his fix in October 2026.

    Published Sep 10, 2026self-described: Hybrid sparse and dense retrieval on self-managed Postgres

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  3. Negative

    Gluber, Hacker News commenter

    Hacker News comment by GluberHacker News comment

    Says pgvector will either trash the caches or take over the CPU of a busy database so that workloads that used to work fine stall, and calls this a Postgres extension issue more than a pgvector one.

    Published Aug 19, 2026self-described: Production RAG on busy Postgres clusters

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  5. Negative

    Aashi Dutt, senior technical content writer at Mem0

    Mem0 blog: How we cut vector search latency by 70xEngineering blog post

    Mem0's HNSW indexes recorded 0 scans while every vector query fell back to an 8-to-14-second sequential scan over a 1,153 GB table, because the planner chose the scan once selective tenant filters entered. Mem0 moved its vectors to turbopuffer and calls staying on pgvector the sensible default under roughly 10 million vectors with simple queries.

    Published Jul 30, 2026self-described: Mem0's production vector search on Aurora PostgreSQL

    Counts toward

    Published on
    Engineering Blogs
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  6. Negative

    itaispiegel (Itai Spiegel), GitHub issue reporter

    GitHub issue by itaispiegelGitHub issue

    A floating pgvector/pgvector:pg17 tag rebuilt from 0.8.2 to 0.8.4 and broke the team's migration-time index build in CI, with an error that 90 MB of memory was required against a 64 MB maintenance_work_mem.

    Published Jul 1, 2026self-described: Migration-time index builds in CI from the official Docker image

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Still applicable

  7. Mixed

    Christian Winther (jippi), previously Staff Engineer at SeatGeek (GitHub profile)

    GitHub issue by jippiGitHub issue

    His team was affected by a Postgres 18 problem and asked for the fix in a release. The maintainer announced 0.8.3 four days later, on June 18, 2026.

    Published Jun 14, 2026self-described: Upgrading to Postgres 18 with pgvector in the stack

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  8. Negative

    TuringNYC, CTO and SVP at an advertising holding company (HN profile)

    Hacker News comment by TuringNYCHacker News comment

    Replying about pgvector on Azure, they say it does not support ultra-wide vectors from the largest embedding models, so their team chops vectors up across fields.

    Published Jun 6, 2026self-described: An enterprise AI creative platform on Azure Postgres

    Counts toward

    Published on
    Developer Communities
    Who
    Technical leadership
    Evidence
    Reported
    Applies to
    Current

  9. Mixed

    Mikhail Nelaev (spyderdfx), Ruby on Rails Developer, abak-press (GitHub profile)

    GitHub issue by spyderdfxGitHub issue

    Nearest-neighbour results on a table of about 20 million binary hashes looked wrong. After he deleted about 15 million duplicates, the index returned the exact match only with hnsw.ef_search raised to 1,000, and the maintainer replied that approximate indexing does not give perfect recall.

    Published Mar 29, 2026self-described: Image-similarity search over binary hashes

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  10. Mixed

    SongYoungUk (R3gardless), Kakao (GitHub profile)

    GitHub issue by R3gardlessGitHub issue

    halfvec cut the HNSW index to 19 GB from 38 GB with vector. A follow-up found halfvec significantly faster when built with GCC 14.2.1, after packaged builds on older compilers lost its speed.

    Published Mar 9, 2026self-described: Benchmarking pgvector with VectorDBBench on Cohere-10M

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  11. Mixed

    alonziv, GitHub issue reporter

    GitHub issue by alonzivGitHub issue

    A query on about 2M rows takes about 8 seconds the first time and 0.3 seconds when repeated, because table reads for the returned rows miss the cache.

    Published Mar 5, 2026self-described: Retrieve-then-rerank on AWS RDS db.m8g.4xlarge

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  12. Mixed

    eb0la, Hacker News commenter

    Hacker News comment by eb0laHacker News comment

    Started with pgvector because the team already knew Postgres and it was easy to hand over to operations. An exact tag match first means 85% of the time they do not need the vector lookup.

    Published Jan 15, 2026self-described: A small RAG system handed over to an operations team

    Counts toward

    Published on
    Developer Communities
    Who
    Small product team
    Evidence
    Reported
    Applies to
    Current

  13. Positive

    juanre, former startup CTO building aweb (HN profile)

    Hacker News comment by juanreHacker News comment

    Quite happy with the performance of their library, which uses PostgreSQL with pgvector, hybrid BM25, multi-query expansion and reranking.

    Published Jan 15, 2026self-described: Retrieval for internal company apps through their llmemory library

    Counts toward

    Published on
    Developer Communities
    Who
    Technical leadership
    Evidence
    Reported
    Applies to
    Current

  14. Positive

    Miro Keimiöniemi (m1r0k3), Clarvo

    Clarvo blog: Optimizing filtered vector queriesEngineering blog post

    Clarvo cut its filtered pgvector queries from up to tens of seconds down to single-digit milliseconds after restructuring them, and lists the practices that did it.

    Published Nov 4, 2025self-described: Clarvo's production recommendation queries

    Counts toward

    Published on
    Engineering Blogs
    Who
    Small product team
    Evidence
    Reported
    Applies to
    Current

  15. Negative

    Daniel Herman (detrin), Machine Learning Engineer, Cisco (GitHub profile)

    GitHub issue by detrinGitHub issue

    Under many high-concurrency scans of 300+ dimension vectors he gets timeouts. He suspects a page lock, and the maintainer replied that removing the lock did not change QPS.

    Published Nov 4, 2025self-described: High-QPS vector search on one Postgres instance

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  16. Positive

    Rafael dos Santos Silva (xfalcox), Discourse (GitHub profile)

    Hacker News comment by xfalcoxHacker News comment

    Discourse runs pgvector in thousands of databases, and it is used in most of the billions of page views Discourse serves.

    Published Nov 3, 2025self-described: Discourse, running pgvector across thousands of hosted databases

    Counts toward

    Published on
    Developer Communities
    Who
    Organization-wide
    Evidence
    Reported
    Applies to
    Current

  17. Positive

    Rafael dos Santos Silva (xfalcox), Discourse (GitHub profile)

    Hacker News comment by xfalcoxHacker News comment

    Indexing binary embeddings and re-ranking at higher precision makes the storage cost of the index 32 times smaller, which they say is what lets Discourse offer it at scale.

    Published Nov 3, 2025self-described: Sizing index storage for Discourse's hosted fleet

    Counts toward

    Published on
    Developer Communities
    Who
    Organization-wide
    Evidence
    Reported
    Applies to
    Current

  18. Negative

    whakim, Halcyon (in the comment)

    Hacker News comment by whakim: Halcyon, with trillions of embeddingsHacker News comment

    Halcyon has trillions of embeddings and found Postgres unsuitable at several orders of magnitude less than that.

    Published Nov 3, 2025self-described: Halcyon, with trillions of embeddings

    Counts toward

    Published on
    Developer Communities
    Who
    Technical leadership
    Evidence
    Reported
    Applies to
    Current

  19. Negative

    Alex Jacobs, practitioner blog

    Alex Jacobs: The case against pgvectorPractitioner blog post

    Building an HNSW index on a few million vectors can consume 10+ GB of RAM, which the blogs he read never mention. He says pgvector is not bad, but real-time inserts with filters at that scale are hard.

    Published Oct 29, 2025self-described: Planning production RAG with real-time inserts and filters on Postgres

    Counts toward

    Published on
    Engineering Blogs
    Who
    Small product team
    Evidence
    Reported
    Applies to
    Current

  20. Mixed

    Ian Myers (igm503), Machine Learning Engineer, Shiftsmart (GitHub profile)

    GitHub issue by igm503GitHub issue

    With a filter of about 1% of the table, the planner kept choosing badly: LIMIT 5 on the HNSW index took 1,885 ms while a parallel sequential scan at LIMIT 40 took 235 ms. The maintainer's advice puts plan control on the team.

    Published Oct 26, 2025self-described: Recency-filtered similarity search over about 3 million vectors

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  21. Mixed

    Jurgen Brandstetter (BrandiATMuhkuh), head of AI and engineering, Howie.systems (HN profile)

    Hacker News comment by BrandiATMuhkuhHacker News comment

    After indexing 1.2 million documents for one customer project, they found Supabase with pgvector needs lots of resources when adding new rows to the index and wishes resources scaled up and down automatically.

    Published Aug 20, 2025self-described: Indexing client documents on Supabase with pgvector

    Counts toward

    Published on
    Developer Communities
    Who
    Technical leadership
    Evidence
    Reported
    Applies to
    Current

  22. Negative

    JoranDox, GitHub issue reporter

    GitHub issue by JoranDoxGitHub issue

    The planner skipped a partial HNSW index. Forcing it was much faster, about 1.5ms against 15 ms, but returned partial results for larger LIMITs.

    Published Jul 23, 2025self-described: Filtered recommendation queries on pgvector

    Counts toward

    Published on
    Developer Communities
    Who
    Product or platform team
    Evidence
    Reported
    Applies to
    Current

  23. Positive

    Ankit Mittal, Engineer at Instacart

    Instacart tech blog: search infrastructure on PostgresEngineering blog post

    In Instacart's production A/B test, pgvector brought a 6% drop in searches with zero results thanks to better recall. It was marginally slower than FAISS for larger retailers.

    Published May 28, 2025self-described: Instacart's search retrieval, moved from FAISS to pgvector

    Counts toward

    Published on
    Engineering Blogs
    Who
    Organization-wide
    Evidence
    Reported
    Applies to
    Current

  24. Mixed

    whakim, first employee at Halcyon (in the comment)

    Hacker News comment by whakim: A RAG startup whose embeddings grew by orders of magnitudeHacker News comment

    The pgvector filtering story is better than a year earlier, but high-cardinality filters still feel like an afterthought. Low-cardinality filters can be solved with partial indexes even at scale.

    Published Mar 9, 2025self-described: A RAG startup whose embeddings grew by orders of magnitude

    Counts toward

    Published on
    Developer Communities
    Who
    Technical leadership
    Evidence
    Reported
    Applies to
    Current

  25. Positive

    rkochanowski, Hacker News commenter

    Hacker News comment by rkochanowskiHacker News comment

    Used halfvec in pgvector on a different project without affecting results.

    Published Jul 4, 2026self-described: Storing embeddings at half precision

    Counts toward

    Published on
    Developer Communities
    Who
    Individual builder
    Evidence
    Reported
    Applies to
    Current

  26. Mixed

    Kyle Ingraham (gomoboo), software developer building Customelon (Kyle Ingraham's own site)

    Hacker News comment by gomobooHacker News comment

    Chose pgvector because it was easy to run, and their corpus is too small for Postgres to elect to use an index, so queries scan the table.

    Published Jan 28, 2026self-described: A niche RAG site on a single server

    Counts toward

    Published on
    Developer Communities
    Who
    Individual builder
    Evidence
    Reported
    Applies to
    Current

Product documentation 7

Pages from pgvector, used to check facts for every group.

  1. README

    pgvector supports two index types for approximate nearest neighbor search: HNSW, which creates a multilayer graph, and IVFFlat, which divides vectors into lists. Adding an index trades some recall for speed. pgvector supports L2 distance, inner product, cosine distance, L1 distance, Hamming distance, and Jaccard distance. Hamming and Jaccard distance apply to binary vectors. pgvector stores single-precision, half-precision, binary, and sparse vectors, with the halfvec, bit and sparsevec types for the last three. Starting with 0.8.0, you can enable iterative index scans, which automatically scan more of the index until enough results are found or the scan reaches `hnsw.max_scan_tuples` or `ivfflat.max_probes`. Iterative scans can use strict or relaxed ordering. By default pgvector performs exact nearest neighbor search, which provides perfect recall. pgvector uses the Postgres write-ahead log (WAL), which allows for replication and point-in-time recovery. pgvector stores your vectors with the rest of your data. A vector can have up to 16,000 dimensions. You can generate and store vectors in one language and query them in another. pgvector is a Postgres extension you compile and install (supports Postgres 13+). You enable it with `CREATE EXTENSION vector` once in each database where you want to use it. The pgvector README lists a client library for Python: pgvector-python. The pgvector README lists a client library for JavaScript and TypeScript: pgvector-node. The pgvector README lists a client library for Java, Kotlin, Groovy and Scala: pgvector-java. On Linux and Mac, you compile and install pgvector (supports Postgres 13+) or install it with Docker, Homebrew, PGXN, APT, Yum, pkg, APK, or conda-forge. On Windows, you build it with nmake or install it with Docker or conda-forge. It comes preinstalled with Postgres.app and many hosted providers. HNSW indexes support `vector` up to 2,000 dimensions, `halfvec` up to 4,000 dimensions, `bit` up to 64,000 dimensions, and `sparsevec` up to 1,000 non-zero elements. IVFFlat indexes support `vector`, `halfvec` and `bit` at the same limits. A vector or half vector can have up to 16,000 dimensions, and a sparse vector up to 16,000 non-zero elements. HNSW query results are limited by the size of the dynamic candidate list (hnsw.ef_search), 40 by default, and NULL vectors (and zero vectors for cosine distance) are not indexed. A non-partitioned table has a limit of 32 TB by default in Postgres. A partitioned table can have thousands of partitions of that size.

    Checked Nov 1, 2026

  2. Neon docs: pgvector

    Neon (Neon's own docs): pgvector is available on every Neon plan with no add-on or paid tier required, and Neon allows installing the previous pgvector version, one version behind the latest it supports.

    Checked Nov 1, 2026

  4. Google Cloud: Cloud SQL extensions

    Cloud SQL for PostgreSQL (Google's own docs) supports pgvector: PostgreSQL versions 13 and later support version 0.8.5, version 12 up to 0.7.4, and version 11 up to 0.5.1.

    Checked Nov 1, 2026

  5. AWS What's New: Amazon RDS supports pgvector

    Amazon RDS for PostgreSQL supports the pgvector extension (AWS What's New, May 3, 2023). It is available on all database instances in Amazon RDS running PostgreSQL 15.2 and higher in all AWS Regions, including the AWS GovCloud (US) Regions.

    Checked Nov 1, 2026

  6. Supabase docs: pgvector

    Supabase (Supabase's own docs): enable pgvector from the Database page in the Dashboard, under Extensions, by searching for `vector`.

    Checked Nov 1, 2026

  7. LICENSE

    pgvector is free under the PostgreSQL License. Its LICENSE grants permission to use, copy, modify, and distribute the software and its documentation for any purpose, without fee, and without a written agreement, provided that the copyright notice and the license text appear in all copies. META.json names the license PostgreSQL. Managed hosting is priced by each host.

    Checked Nov 1, 2026

Source index

Provenance at a glance

Every source GILD read, with the original address and the context used to count it. Open a finding to filter the full cards above. On smaller screens the index lists each source once by name, cohort and date; its full record above carries the address, context and findings.

Provenance for every analyzed Signals source
SourceOriginal URLAuthor and contextWhoPublishedReview dateAppliesFindings
GitHub issue by nlundeeID pg-c102 | Developer Communitiesgithub.com/digdir/genki/issues/186Nikolai Maasø Lunde (nlundee), Digdir (GitHub profile)self-described: Genki, Digdir's shared generative AI platform for the Norwegian public sector, on AzureOrganization-widePublished Oct 29, 2026Assessment reviewed Nov 1, 2026Current
GitHub issue by jitokimID pg-c101 | Developer Communitiesgithub.com/pgvector/pgvector/issues/1025jito (jitokim), Applied AI Engineer, Lbox (GitHub profile)self-described: Hybrid sparse and dense retrieval on self-managed PostgresProduct or platform teamPublished Sep 10, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by GluberID pg-c018 | Developer Communitiesnews.ycombinator.com/item?id=49362150Gluber, Hacker News commenterself-described: Production RAG on busy Postgres clustersProduct or platform teamPublished Aug 19, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by GluberID pg-c019 | Developer Communitiesnews.ycombinator.com/item?id=49362217Gluber, Hacker News commenterself-described: Designing production document RAG systemsProduct or platform teamPublished Aug 19, 2026Assessment reviewed Nov 1, 2026Current
Mem0 blog: How we cut vector search latency by 70xID pg-c032 | Engineering Blogsmem0.ai/blog/how-we-cut-vector-search-latency-by-70xAashi Dutt, senior technical content writer at Mem0self-described: Mem0's production vector search on Aurora PostgreSQLProduct or platform teamPublished Jul 30, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by rkochanowskiID pg-c016 | Developer Communitiesnews.ycombinator.com/item?id=48784282rkochanowski, Hacker News commenterself-described: Storing embeddings at half precisionIndividual builderPublished Jul 4, 2026Assessment reviewed Nov 1, 2026Current
GitHub issue by itaispiegelID pg-c038 | Developer Communitiesgithub.com/pgvector/pgvector/issues/995itaispiegel (Itai Spiegel), GitHub issue reporterself-described: Migration-time index builds in CI from the official Docker imageProduct or platform teamPublished Jul 1, 2026Assessment reviewed Nov 1, 2026Still applicable
GitHub issue by jippiID pg-c104 | Developer Communitiesgithub.com/pgvector/pgvector/issues/987Christian Winther (jippi), previously Staff Engineer at SeatGeek (GitHub profile)self-described: Upgrading to Postgres 18 with pgvector in the stackProduct or platform teamPublished Jun 14, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by TuringNYCID pg-c103 | Developer Communitiesnews.ycombinator.com/item?id=48425549TuringNYC, CTO and SVP at an advertising holding company (HN profile)self-described: An enterprise AI creative platform on Azure PostgresTechnical leadershipPublished Jun 6, 2026Assessment reviewed Nov 1, 2026Current
GitHub issue by spyderdfxID pg-c037 | Developer Communitiesgithub.com/pgvector/pgvector/issues/972Mikhail Nelaev (spyderdfx), Ruby on Rails Developer, abak-press (GitHub profile)self-described: Image-similarity search over binary hashesProduct or platform teamPublished Mar 29, 2026Assessment reviewed Nov 1, 2026Current
GitHub issue by R3gardlessID pg-c036 | Developer Communitiesgithub.com/pgvector/pgvector/issues/967SongYoungUk (R3gardless), Kakao (GitHub profile)self-described: Benchmarking pgvector with VectorDBBench on Cohere-10MProduct or platform teamPublished Mar 9, 2026Assessment reviewed Nov 1, 2026Current
GitHub issue by alonzivID pg-c039 | Developer Communitiesgithub.com/pgvector/pgvector/issues/962alonziv, GitHub issue reporterself-described: Retrieve-then-rerank on AWS RDS db.m8g.4xlargeProduct or platform teamPublished Mar 5, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by gomobooID pg-c014 | Developer Communitiesnews.ycombinator.com/item?id=46803226Kyle Ingraham (gomoboo), software developer building Customelon (Kyle Ingraham's own site)self-described: A niche RAG site on a single serverIndividual builderPublished Jan 28, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by eb0laID pg-c013 | Developer Communitiesnews.ycombinator.com/item?id=46632345eb0la, Hacker News commenterself-described: A small RAG system handed over to an operations teamSmall product teamPublished Jan 15, 2026Assessment reviewed Nov 1, 2026Current
Hacker News comment by juanreID pg-c015 | Developer Communitiesnews.ycombinator.com/item?id=46632355juanre, former startup CTO building aweb (HN profile)self-described: Retrieval for internal company apps through their llmemory libraryTechnical leadershipPublished Jan 15, 2026Assessment reviewed Nov 1, 2026Current
Clarvo blog: Optimizing filtered vector queriesID pg-c012 | Engineering Blogswww.clarvo.ai/blog/optimizing-filtered-vector-queries-from-tens-of-seconds-to-single-digit-milliseconds-in-postgresqlMiro Keimiöniemi (m1r0k3), Clarvoself-described: Clarvo's production recommendation queriesSmall product teamPublished Nov 4, 2025Assessment reviewed Nov 1, 2026Current
GitHub issue by detrinID pg-c106 | Developer Communitiesgithub.com/pgvector/pgvector/issues/918Daniel Herman (detrin), Machine Learning Engineer, Cisco (GitHub profile)self-described: High-QPS vector search on one Postgres instanceProduct or platform teamPublished Nov 4, 2025Assessment reviewed Nov 1, 2026Current
Hacker News comment by xfalcoxID pg-c007 | Developer Communitiesnews.ycombinator.com/item?id=45799162Rafael dos Santos Silva (xfalcox), Discourse (GitHub profile)self-described: Discourse, running pgvector across thousands of hosted databasesOrganization-widePublished Nov 3, 2025Assessment reviewed Nov 1, 2026Current
Hacker News comment by xfalcoxID pg-c008 | Developer Communitiesnews.ycombinator.com/item?id=45805239Rafael dos Santos Silva (xfalcox), Discourse (GitHub profile)self-described: Sizing index storage for Discourse's hosted fleetOrganization-widePublished Nov 3, 2025Assessment reviewed Nov 1, 2026Current
Hacker News comment by whakim: Halcyon, with trillions of embeddingsID pg-c010 | Developer Communitiesnews.ycombinator.com/item?id=45805087whakim, Halcyon (in the comment)self-described: Halcyon, with trillions of embeddingsTechnical leadershipPublished Nov 3, 2025Assessment reviewed Nov 1, 2026Current
Alex Jacobs: The case against pgvectorID pg-c030 | Engineering Blogsalex-jacobs.com/posts/the-case-against-pgvector/Alex Jacobs, practitioner blogself-described: Planning production RAG with real-time inserts and filters on PostgresSmall product teamPublished Oct 29, 2025Assessment reviewed Nov 1, 2026Current
GitHub issue by igm503ID pg-c105 | Developer Communitiesgithub.com/pgvector/pgvector/issues/916Ian Myers (igm503), Machine Learning Engineer, Shiftsmart (GitHub profile)self-described: Recency-filtered similarity search over about 3 million vectorsProduct or platform teamPublished Oct 26, 2025Assessment reviewed Nov 1, 2026Current
Hacker News comment by BrandiATMuhkuhID pg-c017 | Developer Communitiesnews.ycombinator.com/item?id=44958868Jurgen Brandstetter (BrandiATMuhkuh), head of AI and engineering, Howie.systems (HN profile)self-described: Indexing client documents on Supabase with pgvectorTechnical leadershipPublished Aug 20, 2025Assessment reviewed Nov 1, 2026Current
GitHub issue by JoranDoxID pg-c035 | Developer Communitiesgithub.com/pgvector/pgvector/issues/878JoranDox, GitHub issue reporterself-described: Filtered recommendation queries on pgvectorProduct or platform teamPublished Jul 23, 2025Assessment reviewed Nov 1, 2026Current
Instacart tech blog: search infrastructure on PostgresID pg-c031 | Engineering Blogstech.instacart.com/how-instacart-built-a-modern-search-infrastructure-on-postgres-c528fa601d54Ankit Mittal, Engineer at Instacartself-described: Instacart's search retrieval, moved from FAISS to pgvectorOrganization-widePublished May 28, 2025Assessment reviewed Nov 1, 2026Current
Hacker News comment by whakim: A RAG startup whose embeddings grew by orders of magnitudeID pg-c009 | Developer Communitiesnews.ycombinator.com/item?id=43306765whakim, first employee at Halcyon (in the comment)self-described: A RAG startup whose embeddings grew by orders of magnitudeTechnical leadershipPublished Mar 9, 2025Assessment reviewed Nov 1, 2026Current
READMEID pgd-01 | Product Documentationraw.githubusercontent.com/pgvector/pgvector/master/README.mdpgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
Neon docs: pgvectorID pgd-02 | Product Documentationneon.com/docs/extensions/pgvectorpgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
Microsoft Learn: pgvector on Azure Database for PostgreSQLID pgd-03 | Product Documentationlearn.microsoft.com/en-us/azure/postgresql/flexible-server/how-to-use-pgvectorpgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
Google Cloud: Cloud SQL extensionsID pgd-04 | Product Documentationcloud.google.com/sql/docs/postgres/extensionspgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
AWS What's New: Amazon RDS supports pgvectorID pgd-05 | Product Documentationaws.amazon.com/about-aws/whats-new/2023/05/amazon-rds-postgresql-pgvector-ml-model-integration/pgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
Supabase docs: pgvectorID pgd-06 | Product Documentationsupabase.com/docs/guides/database/extensions/pgvectorpgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only
LICENSEID pgd-07 | Product Documentationraw.githubusercontent.com/pgvector/pgvector/master/LICENSEpgvectorOfficialNot counted (documentation)No date statedDocumentation checked Nov 1, 2026CurrentVerification only