# ranking Published articles for ranking. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## 🍔🧠 Pinterest's Fix for the Hardest Problem in ML Infra DevFeed: [🍔🧠 Pinterest's Fix for the Hardest Problem in ML Infra]() Original publisher: [Read original article]() Author: Alexandre Zajac Published: 2026-09-14T15:31:30Z Content type: article Language: en Sources: [Hungry Minds]() Topics: [recommendation systems](), [Feature Engineering](), [Inference](), [Streaming](), [configuration](), [Framework](), [data](), [Low Latency]() Tags: [batch](), [configuration](), [data](), [feature-engineering](), [low-latency](), [ml](), [pinterest](), [ranking](), [recommendation-systems](), [streaming]() ### AI overview Pinterest redesigned its user-sequence platform for ranking, retrieval, and recommendation systems by defining signals once and instantiating them consistently across streaming, batch, and serving workloads. The approach uses Python configuration with validated schemas, a shared execution engine, cooperating streaming and batch paths, and columnar time-partitioned storage to improve freshness, completeness, consistency, and operational efficiency. ### Source excerpt PLUS: OpenAI agents beat math 🧮, Test techniques for agents ⚡, Postgres survival guide 📖 ## Vespa Newsletter, September 2026 DevFeed: [Vespa Newsletter, September 2026]() Original publisher: [Read original article]() Author: Bonnie Chase Published: 2026-09-08T00:00:00Z Content type: news Language: en Sources: [Vespa Blog]() Topics: [ann](), [Latency](), [Provisioning](), [Algorithm](), [telemetry](), [Graphs]() Tags: [2026](), [algorithm](), [cloud](), [features](), [graph](), [latency](), [newsletter](), [product](), [provisioning](), [ranking](), [retrieval](), [september-2026](), [telemetry](), [updates]() ### AI overview The September 2026 Vespa newsletter announces updates including time-constrained ANN search, sub-query ranking support, flexible provisioning, new rank features, and telemetry export. It also introduces Vespa.ai Live, an in-person community meetup focused on retrieval and ranking systems. ### Source excerpt Advances in Vespa include time-constrained ANN search, sub-query ranking support, flexible provisioning, new rank features and telemetry export ## Scaling Conditional Learned Retrieval for Pinterest Home Feed DevFeed: [Scaling Conditional Learned Retrieval for Pinterest Home Feed]() Original publisher: [Read original article]() Author: Pinterest Engineering Published: 2026-08-26T14:01:05Z Content type: article Language: en Sources: [Pinterest Engineering Blog - Medium]() Topics: [recommendation systems](), [Embeddings](), [Machine Learning & Artificial Intelligence]() Tags: [blog](), [embedding](), [eng-culture](), [engineering](), [infrastructure](), [pinner-experience](), [pinterest](), [ranking](), [retrieval]() ### AI overview Pinterest describes scaling Conditional Learned Retrieval for home-feed candidate generation. The approach extends two-tower retrieval with explicit contexts that produce multiple user embeddings for different interests and retrieval use cases. ### Source excerpt Devin Kreuzer | Sr. Machine Learning Engineer; Yichi Wang | Machine Learning Engineer I; Sujan Reddy Ale | Machine Learning Engineer I; Zelun Wang | Sr. Machine Learning Engineer; Hongtao Lin | Sr. Machine Learning Engineer; Piyush Maheshwari | Staff Machine Learning Engineer Pinterest home feed candidate generation is a large-scale User-to-Pin retrieval problem. A common approach is a two-tower model: a user tower encodes the user, an item tower encodes candidate Pins, and approximate nearest neighbor search retrieves Pins close to the user embedding. But Pinterest users often have multiple intentions at once -- planning a renovation, saving recipes, exploring fashion, or organizing travel ideas. A single retrieval embedding can struggle to capture this diversity. Conditional Learned Retrieval, or CLR, extends the two-tower setup by conditioning the user tower on an explicit retrieval context. Instead of producing only one user embedding, CLR can generate condition-aware embeddings that reflect different aspects of a user's interests while still grounding retrieval in the user's overall behavior. Prior Pinterest work studied this formulation in two settings. The RecSys'24 paper: Bootstrapping Conditional Retrieval for User-to-Item Recommendations described how to bootstrap conditional retrieval by constructing training data for (user, condition) -> item retrieval from existing user-item and condition signals, and applied it to interest-based notifications. The KDD'25 paper: Synergizing Implicit and Explicit User Interests: A Multi-Embedding Retrieval Framework at Pinterest placed Conditional Retrieval within a broader multi-embedding retrieval framework for home feed, where explicit interest conditions complement implicit interests extracted from user behavior. In this blog, we describe how CLR evolved from early interest-conditioned retrieval into a broader retrieval system for Pinterest home feed. We focus on three areas: expanding CLR to support more retrieval us ## How Keras 3 Helped Modernise Expedia Group's Lodging Ranking Stack DevFeed: [How Keras 3 Helped Modernise Expedia Group's Lodging Ranking Stack]() Original publisher: [Read original article]() Author: Conor Worthington Published: 2026-08-11T11:01:02Z Content type: article Language: en Sources: [Expedia]() Topics: [Keras](), [recommendation systems](), [Latency](), [Inference Performance](), [Machine learning]() Tags: [data-science](), [features](), [framework](), [improvements](), [inference](), [keras](), [latency](), [llms](), [machine-learning](), [mixture-of-experts](), [model-training](), [neural-networks](), [ranking](), [recommendation-system]() ### AI overview Expedia Group describes modernising its lodging-ranking stack around Keras 3. The changes included rewriting parts of its pipelines, making model training 30% faster and reducing P99 inference latency by two-thirds. The article discusses limitations of the previous Keras 2, eager-training and combined-serving setup, along with the use of newer APIs and accelerator-oriented optimisations. ### Source excerpt Expedia Group Technology -- DataWhat happened when we treated a framework migration as an architecture modernisation -- and cut P99 inference latency by two-thirdsSt Paul's and millennium bridge, London Expedia Group™ has always been a market leader in providing personalised search experiences for travellers. As our ranking models evolved, we saw an opportunity not just to migrate to Keras 3, but to modernise the broader stack around it so we can better serve travellers. This led us to rewrite key parts of our pipelines that made model training 30% faster and cut P99 inference latency by two-thirds. Our main focus in this blog is to discuss our improvements to lodging ranking -- our service which handles users' search requests and returns a personalised property ranking on top of a lightweight candidate generator. This service is frequently retrained to improve customer experience as we get new signals over time. Example of personalised lodging ranking on Expedia search result page for a London search As such, our velocity to make changes needs to be fast, but more importantly we need to be able to build models which are state-of-the-art, enabling customers to easily find and book the most relevant property for their trip. The problem with Keras 2, eager training and combined serving The challenge was not that the old stack was broken. It was that it had gradually become a limiting factor. On the training side, we were carrying dependencies on older components and missing out on modern Keras APIs, new optimisers, new layers and cleaner support for accelerator-oriented training. On the serving side, we had a path that was simple and operationally familiar, but not especially well suited to aggressively optimised inference for heavier ranking architectures. This blockage on utilising new layers and optimisers has become more pronounced as LLMs see blistering amounts of innovation. Naturally, these new features are only available in Keras 3 or as separate backends from ou ## Agentic relevance tuning: Letting LLM agents do the search engineering work DevFeed: [Agentic relevance tuning: Letting LLM agents do the search engineering work]() Original publisher: [Read original article]() Author: Kylie Wagar-Dirks Published: 2026-07-31T15:00:47Z Content type: article Language: en Sources: [OpenSearch]() Topics: [Amazon OpenSearch Service](), [Large Language Model](), [Framework](), [data](), [Deployment]() Tags: [2026](), [agents](), [aws](), [blog](), [data](), [deployment](), [evaluation](), [framework](), [llm-agents](), [opensearch](), [ranking](), [search]() ### AI overview The article introduces Agentic Relevance Tuning (ART), an end-to-end framework that uses specialized LLM-powered agents to automate search relevance improvement in OpenSearch. ART monitors user behavior, proposes ranking changes, runs offline evaluations, and coordinates deployment when improvements are validated. ### Source excerpt At OpenSearchCon Europe 2026, Bobby Mohammed (AWS) and Daniel Wrigley (OpenSource Connections) introduced Agentic Relevance Tuning (ART), a framework that uses specialized LLM agents to fully automate the search relevance lifecycle in OpenSearch. The post Agentic relevance tuning: Letting LLM agents do the search engineering work appeared first on OpenSearch. ## GenRec: Towards LLM-Native Recommendation at Netflix DevFeed: [GenRec: Towards LLM-Native Recommendation at Netflix]() Original publisher: [Read original article]() Author: Netflix Technology Blog Published: 2026-07-30T20:10:15Z Content type: article Language: en Sources: [Netflix](), [Netflix TechBlog - Medium]() Topics: [Netflix](), [recommendation systems](), [Large Language Model](), [LLM Techniques](), [personalization](), [vllm]() Tags: [data](), [feature-engineering](), [genai](), [large-language-models](), [large-language-models-llms](), [llm](), [llms](), [netflix](), [personalization](), [ranking](), [recommendation-system](), [vllm]() ### AI overview Netflix presents GenRec, an LLM-backed recommendation ranker trained on Netflix-specific data and objectives. It converts user histories, item metadata, and context into text, uses a catalog-aware scoring head, aligns recommendations with long-term member value and business goals, and runs in prefill-only mode on Netflix's LLM serving stack. In a large-scale A/B test, GenRec improved short- and long-term online metrics while using fewer labeled examples and input signals than a mature production ranker. ### Source excerpt Authors: Ying Li, Arjun Rao, Shradha Sehgal Introduction Recommendations sit at the heart of the Netflix experience. Our current production models rely on thousands of hand-crafted features over users, items, and interactions, along with specialized architectures for sequence modeling, feature interactions, and multi-task objectives. This stack has evolved over many years to support diverse content types (movies, series, games, live, podcasts) and product surfaces, but its complexity makes it costly to onboard new use cases: adding a content type or surface can require significant feature engineering, architecture change, infrastructure work, and experimentation. At the same time, large language models (LLMs) are changing how we think about recommendation, as shown by recent work such as PLUM, GLIDE, and OneRec-Think. Their broad world knowledge and strong language understanding make it possible to represent user histories and item metadata directly as text, capture rich relationships in a shared semantic space, and steer recommendations via natural-language prompts. However, off-the-shelf LLMs are still far from production-ready recommenders: they often over-recommend globally popular content, hallucinate out-of-catalog items, ignore business constraints, and provide only limited personalization. To address this, we built GenRec, an LLM-backed recommendation ranker that post-trains an internal foundation LLM on Netflix-specific data and objectives. GenRec shows that an LLM-based ranker can match or exceed a mature production system while relying on far fewer labeled examples and input signals. Figure 1: GenRec pipeline. Raw logs of user history, item metadata, and context are transformed via context engineering into natural-language prompts and fed into the GenRec, which runs on vLLM in prefill-only mode and outputs scores for each catalog item, yielding a recommendation ranking. At a high level, GenRec: Verbalizes user histories, item metadata, and context as text ## Access and share AI Gateway leaderboard data DevFeed: [Access and share AI Gateway leaderboard data]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-07-14T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [Artificial Intelligence](), [API](), [CSV](), [inference-providers](), [data]() Tags: [ai](), [ai-gateway](), [api](), [data](), [images](), [inference-providers](), [leaderboard](), [metrics](), [model](), [open](), [ranking](), [time](), [tokens](), [videos]() ### AI overview Vercel has opened the data behind the AI Gateway leaderboards under the CC BY 4.0 license. The data can be downloaded as CSV or queried through an export API, while charts can be shared as branded PNG images. The leaderboards track daily production usage across models, labs, apps, and inference providers. ### Source excerpt We are making the data behind the AI Gateway leaderboards open under the CC BY 4.0 license. You can now download or query the data through the leaderboard-export API endpoint and render any chart as a shareable image. The AI Gateway leaderboards show how AI is used in production, ranking traffic for models, labs, apps, and providers. Data is aggregated daily across trillions of tokens, so you can see what gets adopted and how that changes over time. For deeper analysis, see the July AI Gateway Production Index. What's ranked There are four leaderboards, each with its own metrics: Leaderboard Ranks Metrics Models Individual models Requests, token volume, spend, images or videos generated Labs Model labs Requests, token volume, spend, images or videos generated Apps Opted-in apps built on the AI Gateway Token volume, spend Providers Inference providers Token volume, spend Models and labs can be filtered by modality (text, image, video) and show a daily percentage share over time; apps and providers are aggregated across all modalities and show a ranked top list. Open data The data behind the leaderboards is open, published under Creative Commons Attribution 4.0 (CC BY 4.0). You are free to use, share, and adapt it, including commercially, as long as you give credit, link to the license, and indicate if changes were made. Every chart and ranked list has a download button that exports the current view as a CSV. For programmatic access, use the export endpoint, which returns the same data and is cached for 24 hours: For models and labs, each row is one entity's daily share of a single metric. One response includes rows for requests, tokens, spend, imageCount, and videoCount, so filter on the metric field to pull out the series you want. Share a chart Every chart has a share button that turns the current view into an image. Pick an aspect ratio (landscape, square, or portrait), then download it as a PNG or copy it to your clipboard. The image includes the legend, title, a ## 🗓 This Week In AI Research (1-8 July 26) DevFeed: [🗓 This Week In AI Research (1-8 July 26)]() Original publisher: [Read original article]() Author: Dr. Ashish Bamania Published: 2026-07-12T11:25:32Z Content type: article Language: en Sources: [Into AI]() Topics: [Artificial Intelligence](), [AI Research](), [Large Language Model](), [Transformer](), [qwen](), [grpo](), [Algorithms](), [releases]() Tags: [agentic](), [ai](), [algorithm](), [algorithms](), [grpo](), [llm](), [models](), [performance](), [qwen](), [ranking](), [releases](), [research](), [rl](), [training](), [update]() ### AI overview A weekly roundup of AI research papers and releases highlights findings that reinforcement-learning gains can be concentrated in a single transformer layer and presents LLM-as-a-Verifier, a framework for continuous scoring and ranking of agentic-task solutions. ### Source excerpt The top 10 research papers and AI releases this week (SpaceXAI's Grok 4.5, OpenAI's GPT-Live voice models, Cognition's SWE-1.7, Meta's Muse Spark 1.1, and many more) ## Muse Spark 1.1 is now available on AI Gateway DevFeed: [Muse Spark 1.1 is now available on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-07-09T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [AI Models](), [multimodal](), [long-context](), [Model Context Protocol](), [vercel ai sdk](), [API keys](), [Routing (disambiguation)]() Tags: [agent](), [agentic](), [ai](), [audio](), [context-window](), [image](), [leaderboard](), [model](), [multimodal](), [muse](), [pricing](), [ranking](), [routing](), [sdk](), [spark](), [tokens](), [tool](), [tools]() ### AI overview Muse Spark 1.1 from Meta is now available through Vercel's AI Gateway. It is a multimodal reasoning model with a 1M-token context window for agentic tasks, supporting multiple input types, tool orchestration, MCP servers, custom skills, parallel tool calls, structured output, and search with citations. ### Source excerpt Muse Spark 1.1 from Meta is now available on AI Gateway. It is a multimodal reasoning model with a 1M token context window built for agentic tasks, accepting text, image, video, PDF, and audio inputs. Muse Spark 1.1 plans and orchestrates work across tools and services, operating as a main agent or as a subagent, and it works with new tools, MCP servers, and custom skills without examples. The model supports parallel tool calling, structured output, and built-in search with citations. To use Muse Spark 1.1, set model to meta/muse-spark-1.1 in the AI SDK: AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules, and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests. Try Muse Spark 1.1 in the model playground. Read more ## Grok 4.5 now available on AI Gateway DevFeed: [Grok 4.5 now available on AI Gateway]() Original publisher: [Read original article]() Author: Jerilyn Zheng Published: 2026-07-08T00:00:00Z Content type: release Language: en Sources: [Vercel News]() Topics: [AI Models](), [Claude](), [vercel ai sdk](), [API](), [model-deployment](), [Routing (disambiguation)]() Tags: [ai](), [ai-gateway](), [ai-models](), [api](), [inference](), [leaderboard](), [ranking](), [routing]() ### AI overview Grok 4.5 from SpaceXAI is now available through Vercel AI Gateway. The model accepts text and image inputs, offers configurable reasoning levels, and can be selected through the AI SDK or routing rules. AI Gateway provides unified model access, usage and cost tracking, retries, failover, reporting, API key budgets, and performance optimizations without provider-price markup or platform inference fees. ### Source excerpt Grok 4.5 from SpaceXAI is now available on AI Gateway. Built for coding, knowledge work, and STEM, the model accepts text and image inputs. Grok 4.5 supports low, medium, and high reasoning levels and defaults to high. Set the level with reasoning to balance speed against depth. To use Grok 4.5, set model to xai/grok-4.5 in the AI SDK: You can also set routing rules to switch to Grok 4.5 from other gateway models without touching your code. AI Gateway provides a unified API for calling models, tracking usage and cost, and configuring retries, failover, and performance optimizations for higher-than-provider uptime. It includes built-in custom reporting, Zero Data Retention support, budgets for API keys, routing rules, and more. AI Gateway reflects provider pricing with no markup and does not charge a platform fee on inference, including on Bring Your Own Key (BYOK) requests. Try Grok 4.5 in the model playground. Read more ## Your agent wants to search like a 2010 quant DevFeed: [Your agent wants to search like a 2010 quant]() Original publisher: [Read original article]() Author: Jon Bratseth Published: 2026-07-07T00:00:00Z Content type: opinion Language: en Sources: [Vespa Blog]() Topics: [AI search](), [information retrieval](), [Language models](), [Google Search]() Tags: [agents](), [ai](), [ai-agents](), [bm25](), [genai](), [google-search](), [hybrid-search](), [information-retrieval](), [rag](), [ranking](), [search]() ### AI overview The article argues that AI agents should retrieve information with more control and sophistication than ordinary human search users. It describes a progression from vector retrieval to hybrid search using methods such as BM25 and machine-learned ranking, and presents search as code as a possible next stage. ### Source excerpt The idea of empowering AI agents to retrieve information like a professional is going mainstream. ## Too big to draw, but yet drawable DevFeed: [Too big to draw, but yet drawable]() Original publisher: [Read original article]() Author: Yehonathan Sharvit Published: 2026-07-06T09:00:00Z Content type: article Language: en Sources: [Klipse]() Topics: [ordering](), [structure]() Tags: [aboulafia](), [caustics](), [generative](), [math](), [permutation](), [permutations](), [random](), [ranking](), [recursion](), [visualization]() ### AI overview The third article in a series explains how Aboulafia's tserouf orders all permutations of a word and visualizes them by placing the permutations around a circle and connecting each word to its reversal. Because the full permutation space becomes too large to draw, the article samples chords and shows that they form recurring caustic curves visible at multiple scales. ### Source excerpt Aboulafia's Tserouf - Part 3 of 4 <- Previous: An elegant formulation, inspired by Bill Gates - Next: A wheel, the same forwards and backwards -> ## GenPage: Towards End-to-End Generative Homepage Construction at Netflix DevFeed: [GenPage: Towards End-to-End Generative Homepage Construction at Netflix]() Original publisher: [Read original article]() Author: Netflix Technology Blog Published: 2026-06-29T13:01:02Z Content type: article Language: en Sources: [Netflix](), [Netflix TechBlog - Medium]() Topics: [Netflix](), [recommendation systems](), [Large Language Model](), [Machine learning](), [Reinforcement learning](), [Transformer]() Tags: [ai](), [app](), [diversity](), [engineering](), [feature-engineering](), [generation](), [generative](), [large-language-models](), [machine-learning](), [model](), [netflix](), [optimization](), [ranking](), [recommendation-system](), [reinforcement-learning](), [rl]() ### AI overview This Netflix developer article introduces GenPage, a generative approach that uses a single autoregressive model to construct a personalized homepage by generating recommendation rows, entities, and layout together. It describes replacing a multi-stage recommendation pipeline with end-to-end modeling and using reinforcement learning to optimize whole-page rewards, including interactions such as diversity and the balance between rows. ### Source excerpt Authors: Lequn Wang, Jiangwei Pan, and Linas Baltrunas Figure 1. Autoregressive homepage generation. GenPage builds a Netflix homepage one row or entity at a time, each one conditioned on what's already on the page and the user's context.Introduction The Netflix homepage is the first thing users see when they open the app and the primary way they discover content to enjoy. Almost every part of it is personalized, including which rows appear, which entities show up within those rows, and how everything is arranged on the page. Constructing that homepage is a genuinely hard problem. It is not simply producing one ranked list. The homepage is a structured, two-dimensional layout, made up of recommendation rows and the entities within them. Here, an entity can be a movie, show, game, live event, or other recommendable item. Each choice can affect the value of the others. Traditionally, it is built through a complex, multi-stage pipeline, with separate components for candidate generation and ranking at both the row and entity levels. We saw an opportunity to rethink this design. Large language models have shown that a single generative model can perform diverse tasks just by generating a response to a prompt. Inspired by this prompt-response paradigm, we trained a single generative model to build the homepage by directly answering one question: Given everything we know about this user and this request, what homepage should we generate to maximize user satisfaction? We call this approach GenPage. It treats the user history and request context as the prompt, and autoregressively generates the entire homepage as the response (Figure 1). Unlike most generative recommenders, such as TIGER, HSTU, and OneRec, which generate flat ranked lists, GenPage generates the rows, entities, and layout together. This shift is motivated by several goals: End-to-end modeling. A single transformer model that constructs the page from raw input signals can replace a complex multi-stage recommen ## Distilling Long-Tail User Behavior into Scalable Embeddings for Job Search DevFeed: [Distilling Long-Tail User Behavior into Scalable Embeddings for Job Search]() Original publisher: [Read original article]() Author: Marsan Ma Published: 2026-06-03T23:28:10Z Content type: article Language: en Sources: [Indeed]() Topics: [recommendation systems](), [Embeddings](), [Deep learning](), [Latency](), [Job]() Tags: [big-data](), [data-science](), [deep-learning](), [embeddings](), [engineering](), [feature-store](), [job-search](), [latency](), [machine-learning](), [performance](), [ranking]() ### AI overview Indeed describes a user behavior modeling system for job search that learns from long-term user histories offline, distills them into fixed-length embeddings, and serves them through a feature store for use by online ranking and recommendation models. The approach is designed to preserve rich behavioral signals while meeting latency and cost constraints. ### Source excerpt Authors : Marsan Ma, Nikhil Lopes, Raj Amrit, Hong Lu, Dipankar Biswas, Trent KyonoLeadership: Iris Wang, Madhu Kurup Recommendation and ranking systems power many of the most important experiences on large internet platforms. Yet the models that run in production are rarely the largest models we can train. They are usually compact, latency-sensitive supervised models [...] ## Re-autoresearching MSMARCO BM25, on Vespa DevFeed: [Re-autoresearching MSMARCO BM25, on Vespa]() Original publisher: [Read original article]() Author: andreer thomas Published: 2026-05-29T00:00:00Z Content type: article Language: en Sources: [Vespa Blog]() Topics: [Benchmark](), [dataset](), [Python](), [Large Language Model](), [generalization in machine learning](), [pandas](), [Google Search]() Tags: [benchmark](), [bm25](), [dataset](), [evaluation](), [experiment](), [google-search](), [information-retrieval](), [openai](), [pandas](), [python](), [ranking](), [retrieval](), [zero-shot]() ### AI overview This article reproduces an MSMARCO BM25 autoresearch experiment in Vespa. It compares LLM-driven Python reranking with an approach restricted to existing Vespa rank features and reports a comparable improvement on a 650,000-passage subset, with better generalization to the full dataset. ### Source excerpt BM25 is having a moment. We reproduce Doug Turnbull's MSMARCO autoresearch experiment in Vespa and get a comparable MRR@10 lift from existing rank features -- with twice the generalization to full MSMARCO. ## Vespa Newsletter, May 2026 DevFeed: [Vespa Newsletter, May 2026]() Original publisher: [Read original article]() Author: Bonnie Chase Published: 2026-05-27T00:00:00Z Content type: news Language: en Sources: [Vespa Blog]() Topics: [AI search](), [Artificial Intelligence](), [Cloud](), [code productivity](), [configuration]() Tags: [ai](), [cloud](), [configuration](), [developer-productivity](), [embedding](), [newsletter](), [ranking](), [search](), [vector-search]() ### AI overview The May 2026 Vespa newsletter announces updates for retrieval and ranking systems, including improved ranking flexibility, embedding integrations with Voyage AI, OpenAI, and Mistral AI, Vespa Cloud dashboards and backups, maintenance controls, custom resource tags, agent skills, and new query and array-field capabilities. ### Source excerpt Advances in Vespa include finer control over deployments, smarter ranking, richer embedding integrations, and more scalable vector search. ## Introducing the Ettin Reranker Family DevFeed: [Introducing the Ettin Reranker Family]() Original publisher: [Read original article]() Author: Tom Aarsen Published: 2026-05-19T00:00:00Z Content type: article Language: en Sources: [Hugging Face - Blog]() Topics: [sentence-transformers](), [Transformer](), [long-context](), [Agent Skill]() Tags: [ai-coding](), [community](), [long-context](), [nlp](), [open-source](), [ranking](), [research](), [retrieval](), [sentence-transformers]() ### AI overview The article introduces six Sentence Transformers CrossEncoder rerankers built on Ettin ModernBERT encoders. It explains their distillation-based training, retrieval-then-rerank usage, quality and cost trade-offs, and support for up to 8K tokens of context. ### Source excerpt We're on a journey to advance and democratize artificial intelligence through open source and open science. ## ML based ranking using Nrtsearch DevFeed: [ML based ranking using Nrtsearch]() Original publisher: [Read original article]() Author: Mohammad Mohtasham (Software Engineer); Tao Yu (Software Engineer) Published: 2026-05-11T00:00:00Z Content type: tutorial Language: en Sources: [Yelp]() Topics: [Machine Learning & Artificial Intelligence](), [Inference](), [AI Inference](), [model-deployment](), [backends](), [Open Source]() Tags: [blog-post](), [inference](), [latency](), [machine-learning](), [ml](), [models](), [network](), [open-source](), [performance](), [pipeline](), [ranking](), [software]() ### AI overview Yelp extended its Lucene-based Nrtsearch engine with an Inference Plugin that embeds machine-learning ranking directly in the search layer. The article explains the ranking workflow, including model configuration and loading, and describes how co-locating feature storage and inference reduces network transfer, serialization overhead, and latency compared with a standalone inference service. ### Source excerpt We've extended Nrtsearch with the Inference Plugin, which embeds ML-based ranking directly in the search layer -- eliminating the need for a standalone scoring service. We use Nrtsearch (read more information on the blog post), a Lucene-based open-source search engine built by Yelp, to power a variety of applications such as business search, reviews search, ad delivery and photo search. In this blog post, we give a high-level overview of the Machine Learning (ML) based scoring workflow in Nrtsearch. We'll show how ML models are configured and loaded, and how different applications use custom business logic to develop, test, and... ## How Whatnot's Hourly ML Feature Pipeline Safely Handled a Failed Update DevFeed: [How Whatnot's Hourly ML Feature Pipeline Safely Handled a Failed Update]() Original publisher: [Read original article]() Author: Whatnot Engineering Published: 2026-05-07T16:01:01Z Content type: article Language: en Sources: [Whatnot Engineering]() Topics: [Machine learning](), [recommendation systems](), [datasets](), [data](), [real-time](), [Processes]() Tags: [ai](), [data](), [data-engineering](), [feature](), [feature-engineering](), [feature-store](), [machine-learning](), [ml](), [pipeline](), [ranking](), [real-time](), [recommendation-system](), [recommendations](), [run]() ### AI overview Whatnot describes how its hourly ML feature pipeline handled a bad pull request that caused a feature group to produce zero rows. The online feature store served last-known-good features for up to two days, anomaly monitors detected the row-count drop, and the team reverted the change before the freshness gap became significant. The post shares lessons about safe degradation and operating frequently updated recommendation features. ### Source excerpt David Kawashima, Miguel Fernandez-Montes Cuberta, Ford Bohrmann | Discovery A bad PR went out just as our hourly feature pipeline was about to run. The pipeline was responsible for recomputing fresh user behavior signals used in livestream ranking on the home feed. These features capture recent user intent: watches, bids, orders aggregated across users, sellers, and categories. A missed feature update means that the system stops responding to user signals immediately. This PR release caused one of our feature groups to suddenly produce zero rows ... but nothing broke! Even though a critical feature group for the model stopped updating, the online feature store continued serving the last-known-good features, buffered by a 2-day TTL and nothing broke explicitly. Our anomaly monitors went off immediately, flagging the drop in row counts. The team identified the bad change and reverted it before the freshness gap grew large enough to matter. No corrupted data reached production. The system absorbed the failure and degraded safely. This graceful degradation didn't happen by accident; it was the result of hardened processes from years of lived experience, seeing the way these systems can degrade in obvious or subtle ways. In this post, we share some of those lessons learned. The system: hourly feature pipeline Whatnot is a highly dynamic marketplace where buyers' preferences change at a moment's notice. A buyer might research a laptop, join a football-card break a few hours later, and buy chocolate before logging off for the night. Daily feature updates capture long-term preferences but miss intra-day shifts. To keep up with these rapidly changing buyer interests (in 2025) we moved from daily to hourly updates for features that power our ML feed ranking models. Fresher features are a critical component for capturing the live nature of the Whatnot marketplace, where the inventory available to purchase changes in real time. Moving from daily to hourly updates was a meaningful ## Empowering Carrot Ads with Domain Adaptive Learning DevFeed: [Empowering Carrot Ads with Domain Adaptive Learning]() Original publisher: [Read original article]() Author: Xiyu Wang Published: 2026-05-04T19:11:17Z Content type: article Language: en Sources: [Instacart]() Topics: [recommendation systems](), [Artificial Intelligence](), [data](), [Optimization](), [Website]() Tags: [advertising](), [ai](), [data](), [domain](), [machine-learning](), [optimization](), [ranking](), [retail]() ### AI overview The article explains how Instacart applies Domain Adaptive Learning to Carrot Ads to address the cold-start problem for new retail partner websites. It describes transferring knowledge from Instacart Marketplace data to target domains to improve click-through-rate prediction and ad ranking with limited partner-specific interaction data. ### Source excerpt Authors: Trey Zhong, Xiyu Wang Contributors: Joseph Haraldson, Sharad Gupta, Sarah Lamacchia Introduction Carrot Ads is Instacart's omnichannel retail media solution that allows retailer partners to build and scale their own advertising businesses on either their owned-and-operated (O&O) websites and apps or their whitelabel Storefront hosted by Instacart. Carrot Ads empowers retailers and CPG brands to accelerate revenue, while improving the customer experience, engagement and Ads return on investment. It features enterprise-grade infrastructure, AI-powered optimization, years of proprietary first-party data and flexibility to choose from retailer-sourced Ads demand, Instacart-sourced demand from 7,500+ CPG brands, or both. However, onboarding a new partner onto Carrot Ads introduces a key challenge: the 'cold start' problem, where limited historical interactions make it difficult to predict user behavior accurately. To serve performant ads, our systems rely on predicting a user's Click-Through Rate (CTR) to generate a ranking score. On the Instacart Marketplace, we have billions of historical signals to train a model to do so. But when a partner launches a new ads experience on their O&O e-commerce site, there is often little to no interaction history for that property, so training an accurate model becomes challenging. User behavior can vary dramatically between websites -- for example, browsing patterns on a grocery site differ from those on a pet supply or electronics site. Training a model from scratch for a new domain is data hungry. Conversely, directly deploying Instacart's existing Marketplace model often fails to capture the nuances of the partner's specific inventory and user base. To address this, we developed a Domain Adaptive Learning approach that transfers knowledge from Instacart's data-rich environment to new partner environments. By treating the Instacart Marketplace as a source domain and the partner's website as a target domain, we can transfer ## Introducing the OSSCAR Index DevFeed: [Introducing the OSSCAR Index]() Original publisher: [Read original article]() Author: Prashant Sridharan Published: 2026-04-28T07:00:00Z Content type: article Language: en Sources: [Supabase Blog]() Topics: [Open Source](), [data](), [Code](), [OpenClaw]() Tags: [2026](), [agent](), [code](), [data](), [open-source](), [openclaw](), [ranking]() ### AI overview The OSSCAR Index is a quarterly ranking of the fastest-growing open source organizations, using transparent and reproducible measurements. Its first edition covers Q1 2026, and the site, data, and scoring code are open source. ### Source excerpt Announcing the OSSCAR Index: a quarterly ranking of the fastest-growing open source organizations. The site, the data, and the scoring code are all open source. ## Scaling a Vespa Application: Feeding Fast and Furiously DevFeed: [Scaling a Vespa Application: Feeding Fast and Furiously]() Original publisher: [Read original article]() Author: Kai Borgen Published: 2026-04-28T00:00:00Z Content type: tutorial Language: en Sources: [Vespa Blog]() Topics: [AI search](), [Deployment](), [datasets](), [information retrieval](), [Cloud](), [Homebrew](), [Command-line interface](), [onnx](), [optimum](), [XML]() Tags: [ai](), [ai-search](), [cli](), [cloud](), [dataset](), [genai](), [how-to](), [information-retrieval](), [install](), [metrics](), [onnx](), [optimum](), [performance](), [rag](), [ranking](), [tensors](), [tutorial]() ### AI overview This tutorial demonstrates how to scale a Vespa application while feeding the full MS_marco passages dataset. It covers preparing the dataset, configuring access, deploying a sample application, and using scaling and metrics to improve feed throughput and performance. ### Source excerpt A tutorial on how to scale the resources in a Vespa application to increase feed throughput. Using the metrics dashboard for informed and optimised scaling. ## Improving Embedding-Based Candidate Generation for Recommender Systems with a Two-Tower Model DevFeed: [Improving Embedding-Based Candidate Generation for Recommender Systems with a Two-Tower Model]() Original publisher: [Read original article]() Author: Ahmad Khan Published: 2026-03-18T21:04:58Z Content type: article Language: en Sources: [Glassdoor Engineering]() Topics: [recommendation systems](), [Machine learning](), [Embeddings](), [Neural Network](), [data](), [ordering]() Tags: [ai](), [data](), [embedding](), [machine-learning](), [model-training](), [neural-network](), [ordering](), [ranking](), [recommender-systems](), [self-supervised-learning](), [two-tower-model]() ### AI overview Glassdoor describes a two-tower neural network for candidate generation in its Community feed recommendation system. The model learns user and post embeddings from interaction data to retrieve potentially relevant posts before filtering, ranking, and ordering. ### Source excerpt Overview At Glassdoor (part of Indeed), millions of professionals come to learn about workplaces, ask career questions, and share experiences through our Community platform. When users open the Community Home Feed, they expect to see posts that are relevant to their interests, industry, and career stage. However, generating that feed is not a simple task. At any moment, thousands of new posts across hundreds of community feeds compete for attention. From discussions about interview experiences to workplace culture questions and industry advice, the platform contains a diverse and constantly evolving stream of content. The challenge is to quickly identify which of these posts are most relevant for each user. To address this problem, we built a multi-stage machine learning recommendation system that efficiently sources, filters, and ranks posts before they appear in the user's feed. In a previous post, Inside Glassdoor's Multi-Stage Recommendation System, we described the architecture of this system and how the different stages work together to deliver personalized content. Snapshot of the Glassdoor Community Home Feed At a high level, the system consists of four stages: Candidate Generation -- retrieves a large set of potentially relevant posts Filtering -- removes posts that are undesirable, duplicated, or already seen Ranking -- predicts which posts a user is most likely to engage with Ordering -- applies heuristics to maintain diversity and freshness in the feed Each stage progressively narrows the set of posts until only the most relevant posts remain. 4 Stage ML RecSys Design In this blog post, we focus on the Candidate Generation stage, which is responsible for retrieving an initial pool of potentially relevant posts from a very large pool of posts across many different feeds. For Candidate Generation, we developed a Two-Tower Neural Network Model that learns embeddings for users and posts based on real interaction data. In the rest of this blog post, we will walk ## Asymmetric Retrieval: Spend on Docs, Embed your Queries for Free DevFeed: [Asymmetric Retrieval: Spend on Docs, Embed your Queries for Free]() Original publisher: [Read original article]() Author: thomas bjorncs Published: 2026-03-10T00:00:00Z Content type: article Language: en Sources: [Vespa Blog]() Topics: [Embeddings](), [Latency](), [API]() Tags: [ai](), [api](), [container](), [cost](), [embedding](), [genai](), [latency](), [local](), [model](), [models](), [quality](), [rag](), [ranking](), [vector](), [voyage-ai]() ### AI overview The article presents asymmetric retrieval using Voyage AI models with Vespa. Documents are embedded once with a high-quality, API-based model, while frequently submitted queries are embedded with a small model running locally. Because the models share a compatible vector space, this approach can eliminate recurring query-embedding API costs, reduce latency, and allow query models to be upgraded independently. ### Source excerpt Documents are embedded once -- worth the spend for maximum quality. Queries hit you on every request. This is what drives your cost at scale. Asymmetric retrieval with Voyage AI and Vespa. Real numbers, real config. [Next page]()