<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>Chris Simpson</title>
        <link>https://chrissimpson.co.uk</link>
        <description>Software developer — writing on software, search, and infrastructure.</description>
        <lastBuildDate>Thu, 28 May 2026 00:27:16 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <image>
            <title>Chris Simpson</title>
            <url>https://chrissimpson.co.uk/favicon.ico</url>
            <link>https://chrissimpson.co.uk</link>
        </image>
        <copyright>All rights reserved 2026</copyright>
        <item>
            <title><![CDATA[Asymmetric query quantization in DiskBBQ]]></title>
            <link>https://chrissimpson.co.uk/asymmetric-query-quantization-in-diskbbq.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/asymmetric-query-quantization-in-diskbbq.html</guid>
            <pubDate>Wed, 27 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Ben Trent and Thomas Veasey shipped another DiskBBQ optimisation today: quantising queries against coarser parent centroids instead of per-document ones. 5x off the quantisation stack with no meaningful recall loss. The insight underneath is that the query path and the document path don’t have to be treated symmetrically.]]></description>
            <content:encoded><![CDATA[Ben Trent and Thomas Veasey shipped another DiskBBQ optimisation today: quantising queries against coarser parent centroids instead of per-document ones. 5x off the quantisation stack with no meaningful recall loss. The insight underneath is that the query path and the document path don’t have to be treated symmetrically.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[SMART: late interaction without retraining]]></title>
            <link>https://chrissimpson.co.uk/smart-late-interaction-without-retraining.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/smart-late-interaction-without-retraining.html</guid>
            <pubDate>Tue, 26 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A new paper out this week shows the per-token hidden states of off-the-shelf single-vector embedders already carry the information needed for ColBERT-style MaxSim - and you can wire it in at inference time, without retraining. The late-interaction deployment barrier I most underestimated just dropped.]]></description>
            <content:encoded><![CDATA[A new paper out this week shows the per-token hidden states of off-the-shelf single-vector embedders already carry the information needed for ColBERT-style MaxSim - and you can wire it in at inference time, without retraining. The late-interaction deployment barrier I most underestimated just dropped.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Faster similar-document search in Elasticsearch 9.4]]></title>
            <link>https://chrissimpson.co.uk/faster-similar-document-search-in-elasticsearch-9-4.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/faster-similar-document-search-in-elasticsearch-9-4.html</guid>
            <pubDate>Thu, 21 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Elasticsearch 9.4 adds query_vector_builder.lookup - a tiny API addition that collapses a two-request vector search into one and runs better than 3x faster. A small change with a big impact, and a look at where that ratio actually comes from.]]></description>
            <content:encoded><![CDATA[Elasticsearch 9.4 adds query_vector_builder.lookup - a tiny API addition that collapses a two-request vector search into one and runs better than 3x faster. A small change with a big impact, and a look at where that ratio actually comes from.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[SID-1: Train the loop, keep the index]]></title>
            <link>https://chrissimpson.co.uk/sid-1-train-the-loop-keep-the-index.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/sid-1-train-the-loop-keep-the-index.html</guid>
            <pubDate>Wed, 20 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[SID AI’s SID-1 is the first retrieval model trained end-to-end with RL. Some observations through a search-and-IR lens: the middle of the retrieval pipeline collapses into one trained model, the NDCG reward gets deliberately bent toward recall, and the agentic-retrieval loop becomes a subagent you hand to a larger system.]]></description>
            <content:encoded><![CDATA[SID AI’s SID-1 is the first retrieval model trained end-to-end with RL. Some observations through a search-and-IR lens: the middle of the retrieval pipeline collapses into one trained model, the NDCG reward gets deliberately bent toward recall, and the agentic-retrieval loop becomes a subagent you hand to a larger system.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[The harness is mostly retrieval]]></title>
            <link>https://chrissimpson.co.uk/the-harness-is-mostly-retrieval.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/the-harness-is-mostly-retrieval.html</guid>
            <pubDate>Mon, 18 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Laurie Voss says applied-AI iteration has moved off the model and into "the harness". He’s right - and once you strip the new vocabulary, the harness is mostly a retrieval system.]]></description>
            <content:encoded><![CDATA[Laurie Voss says applied-AI iteration has moved off the model and into "the harness". He’s right - and once you strip the new vocabulary, the harness is mostly a retrieval system.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[xAI algorithm through a search lens]]></title>
            <link>https://chrissimpson.co.uk/xai-algorithm-through-a-search-lens.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/xai-algorithm-through-a-search-lens.html</guid>
            <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[xAI open-sourced the For You feed algorithm today. Three observations through a search-and-IR lens: two-tower’s quiet dominance, the retrieve/rank split surviving the bitter lesson, and recsys converging with search.]]></description>
            <content:encoded><![CDATA[xAI open-sourced the For You feed algorithm today. Three observations through a search-and-IR lens: two-tower’s quiet dominance, the retrieve/rank split surviving the bitter lesson, and recsys converging with search.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Could TurboQuant Unlock Late Interaction Retrieval?]]></title>
            <link>https://chrissimpson.co.uk/could-turboquant-unlock-late-interaction-retrieval.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/could-turboquant-unlock-late-interaction-retrieval.html</guid>
            <pubDate>Fri, 15 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[TurboQuant landed as a KV cache result, but the more interesting application might be ColBERT-style late interaction. Here’s the case, and the open questions.]]></description>
            <content:encoded><![CDATA[TurboQuant landed as a KV cache result, but the more interesting application might be ColBERT-style late interaction. Here’s the case, and the open questions.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[What to make of TurboQuant]]></title>
            <link>https://chrissimpson.co.uk/what-to-make-of-turboquant.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/what-to-make-of-turboquant.html</guid>
            <pubDate>Thu, 14 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A new quantisation method out of Google Research is making the rounds. Qdrant shipped it. Elastic ran the benchmarks and politely declined. Both responses tell you something useful.]]></description>
            <content:encoded><![CDATA[A new quantisation method out of Google Research is making the rounds. Qdrant shipped it. Elastic ran the benchmarks and politely declined. Both responses tell you something useful.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[The pattern goes all the way down]]></title>
            <link>https://chrissimpson.co.uk/the-pattern-goes-all-the-way-down.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/the-pattern-goes-all-the-way-down.html</guid>
            <pubDate>Wed, 13 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[DiskBBQ’s new filtered-search optimisation is the same architectural move I wrote about last week, applied one layer deeper. The pattern is fractal - and that’s what makes it useful.]]></description>
            <content:encoded><![CDATA[DiskBBQ’s new filtered-search optimisation is the same architectural move I wrote about last week, applied one layer deeper. The pattern is fractal - and that’s what makes it useful.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Everything has changed in search. Nothing has changed in search.]]></title>
            <link>https://chrissimpson.co.uk/everything-and-nothing-has-changed-in-search.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/everything-and-nothing-has-changed-in-search.html</guid>
            <pubDate>Tue, 05 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Agentic search looks like a clean break from classical IR. Look closer and the architectural instincts are the ones backend engineers have used for decades - the components are new, the rules are not.]]></description>
            <content:encoded><![CDATA[Agentic search looks like a clean break from classical IR. Look closer and the architectural instincts are the ones backend engineers have used for decades - the components are new, the rules are not.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[A primer on late interaction]]></title>
            <link>https://chrissimpson.co.uk/a-primer-on-late-interaction.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/a-primer-on-late-interaction.html</guid>
            <pubDate>Fri, 02 May 2025 00:00:00 GMT</pubDate>
            <description><![CDATA[How ColBERT-style token-level matching fits between single-vector dense retrieval and cross-encoders, why MaxSim is the clever bit, and what the storage tax actually looks like in practice.]]></description>
            <content:encoded><![CDATA[How ColBERT-style token-level matching fits between single-vector dense retrieval and cross-encoders, why MaxSim is the clever bit, and what the storage tax actually looks like in practice.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Today, Vimeo goes public]]></title>
            <link>https://chrissimpson.co.uk/vimeo-goes-public.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/vimeo-goes-public.html</guid>
            <pubDate>Tue, 25 May 2021 00:00:00 GMT</pubDate>
            <description><![CDATA[Vimeo spins out from IAC and begins trading on Nasdaq under the ticker VMEO.]]></description>
            <content:encoded><![CDATA[Vimeo spins out from IAC and begins trading on Nasdaq under the ticker VMEO.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[FOSDEM 2015]]></title>
            <link>https://chrissimpson.co.uk/fosdem-2015.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/fosdem-2015.html</guid>
            <pubDate>Mon, 02 Feb 2015 00:00:00 GMT</pubDate>
            <description><![CDATA[Another year at FOSDEM — Vimeo's open source talk, the dedicated Open Source Search track, and a closing keynote from a Mars One astronaut candidate.]]></description>
            <content:encoded><![CDATA[Another year at FOSDEM — Vimeo's open source talk, the dedicated Open Source Search track, and a closing keynote from a Mars One astronaut candidate.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[I've Joined Vimeo]]></title>
            <link>https://chrissimpson.co.uk/ive-joined-vimeo.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/ive-joined-vimeo.html</guid>
            <pubDate>Mon, 07 Apr 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[Joining the Vimeo team to work on the search platform after an amazing run at DueDil.]]></description>
            <content:encoded><![CDATA[Joining the Vimeo team to work on the search platform after an amazing run at DueDil.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[DueDil raises further $17 million of funding to fuel growth and expansion]]></title>
            <link>https://chrissimpson.co.uk/duedil-raises-17m-series-b-funding.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/duedil-raises-17m-series-b-funding.html</guid>
            <pubDate>Mon, 03 Mar 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[DueDil's Series B: a $17m round led by Oak Investment Partners, bringing total funding to $22m and accelerating expansion to new geographies.]]></description>
            <content:encoded><![CDATA[DueDil's Series B: a $17m round led by Oak Investment Partners, bringing total funding to $22m and accelerating expansion to new geographies.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Elasticsearch 1.0 launched: An overview]]></title>
            <link>https://chrissimpson.co.uk/elasticsearch-v1-launched.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/elasticsearch-v1-launched.html</guid>
            <pubDate>Wed, 12 Feb 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[A run-down of the headline features in Elasticsearch 1.0 — Snapshot/Restore, the cat API, the redesigned percolator, and the new Aggregations framework.]]></description>
            <content:encoded><![CDATA[A run-down of the headline features in Elasticsearch 1.0 — Snapshot/Restore, the cat API, the redesigned percolator, and the new Aggregations framework.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[FOSDEM 2014: a retrospective]]></title>
            <link>https://chrissimpson.co.uk/fosdem-2014-retrospective.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/fosdem-2014-retrospective.html</guid>
            <pubDate>Sun, 02 Feb 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[A weekend in Brussels at FOSDEM — Elasticsearch 1.0 ahead of launch, plus PostgreSQL JSON, Redis, MongoDB, and YARN talks.]]></description>
            <content:encoded><![CDATA[A weekend in Brussels at FOSDEM — Elasticsearch 1.0 ahead of launch, plus PostgreSQL JSON, Redis, MongoDB, and YARN talks.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Elasticsearch Marvel: Monitor and Manage your Elasticsearch cluster]]></title>
            <link>https://chrissimpson.co.uk/elasticsearch-marvel-monitor-and-manage-your-cluster.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/elasticsearch-marvel-monitor-and-manage-your-cluster.html</guid>
            <pubDate>Tue, 28 Jan 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[Elasticsearch's new Marvel monitoring dashboard — built on Kibana and Sense — surfaces cluster health and lets you query the REST API live.]]></description>
            <content:encoded><![CDATA[Elasticsearch's new Marvel monitoring dashboard — built on Kibana and Sense — surfaces cluster health and lets you query the REST API live.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Elasticsearch Snapshot Restore Overview]]></title>
            <link>https://chrissimpson.co.uk/elasticsearch-snapshot-restore-api.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/elasticsearch-snapshot-restore-api.html</guid>
            <pubDate>Sun, 05 Jan 2014 00:00:00 GMT</pubDate>
            <description><![CDATA[A walkthrough of the new Snapshot/Restore API arriving in Elasticsearch 1.0 — incremental backups for your cluster via a simple REST endpoint.]]></description>
            <content:encoded><![CDATA[A walkthrough of the new Snapshot/Restore API arriving in Elasticsearch 1.0 — incremental backups for your cluster via a simple REST endpoint.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Elasticsearch Aggregations Overview]]></title>
            <link>https://chrissimpson.co.uk/elasticsearch-aggregations-overview.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/elasticsearch-aggregations-overview.html</guid>
            <pubDate>Wed, 18 Dec 2013 00:00:00 GMT</pubDate>
            <description><![CDATA[A look at the new Aggregations framework arriving in Elasticsearch 1.0 — multi-level, nested calculations that go far beyond what Facets could do.]]></description>
            <content:encoded><![CDATA[A look at the new Aggregations framework arriving in Elasticsearch 1.0 — multi-level, nested calculations that go far beyond what Facets could do.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[London Elasticsearch User Group Presentation]]></title>
            <link>https://chrissimpson.co.uk/london-elasticsearch-usergroup-presentation.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/london-elasticsearch-usergroup-presentation.html</guid>
            <pubDate>Tue, 10 Sep 2013 00:00:00 GMT</pubDate>
            <description><![CDATA[A talk at the London Elasticsearch meetup on how DueDil uses Elasticsearch — bulk indexing, and using Facets to add depth to search.]]></description>
            <content:encoded><![CDATA[A talk at the London Elasticsearch meetup on how DueDil uses Elasticsearch — bulk indexing, and using Facets to add depth to search.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Elasticsearch "Yellow" cluster status explained]]></title>
            <link>https://chrissimpson.co.uk/elasticsearch-yellow-cluster-status-explained.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/elasticsearch-yellow-cluster-status-explained.html</guid>
            <pubDate>Sat, 20 Apr 2013 00:00:00 GMT</pubDate>
            <description><![CDATA[What yellow status actually means, and how the Cluster Health API reports primary vs replica shard allocation.]]></description>
            <content:encoded><![CDATA[What yellow status actually means, and how the Cluster Health API reports primary vs replica shard allocation.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[DueDil completes Series A funding and announces $5m in new investment]]></title>
            <link>https://chrissimpson.co.uk/duedil-series-a-funding-announcement.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/duedil-series-a-funding-announcement.html</guid>
            <pubDate>Thu, 11 Apr 2013 00:00:00 GMT</pubDate>
            <description><![CDATA[DueDil's $5m Series A round, led by Notion Capital and Oak Investment Partners — funding new territories and additional data sets.]]></description>
            <content:encoded><![CDATA[DueDil's $5m Series A round, led by Notion Capital and Oak Investment Partners — funding new territories and additional data sets.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[Using Elasticsearch on Amazon EC2]]></title>
            <link>https://chrissimpson.co.uk/using-elasticsearch-on-amazon-ec2.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/using-elasticsearch-on-amazon-ec2.html</guid>
            <pubDate>Sat, 26 Jan 2013 00:00:00 GMT</pubDate>
            <description><![CDATA[Setting up an Elasticsearch cluster on EC2: installing the AWS cloud plugin, configuring discovery, and watching nodes find each other.]]></description>
            <content:encoded><![CDATA[Setting up an Elasticsearch cluster on EC2: installing the AWS cloud plugin, configuring discovery, and watching nodes find each other.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
        <item>
            <title><![CDATA[DueDil: Trust the data]]></title>
            <link>https://chrissimpson.co.uk/duedil-trust-the-data-video.html</link>
            <guid isPermaLink="false">https://chrissimpson.co.uk/duedil-trust-the-data-video.html</guid>
            <pubDate>Sat, 29 Dec 2012 00:00:00 GMT</pubDate>
            <description><![CDATA[DueDil's intro video — a quick look at what we're building.]]></description>
            <content:encoded><![CDATA[DueDil's intro video — a quick look at what we're building.]]></content:encoded>
            <author>chris@chrissimpson.co.uk (Chris Simpson)</author>
        </item>
    </channel>
</rss>