# fasttext Published articles for fasttext. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## SHIFTing Languages in Multilingual RAG DevFeed: [SHIFTing Languages in Multilingual RAG]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-09-16T07:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Retrieval Augmented Generation (RAG)](), [information retrieval](), [Embeddings](), [AI search](), [qwen](), [qwen3]() Tags: [approximate-nearest-neighbor-search](), [bert](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [information-retrieval](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [qwen](), [qwen3](), [rag](), [recommender-system](), [saas](), [search](), [semantic-search](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This article examines language bias in multilingual retrieval-augmented generation systems. It explains how multilingual embedding models can favor documents written in the query language, making relevant answers in other languages harder to retrieve, and discusses translation, per-language search, and larger embedding models as possible responses. ### Source excerpt If you speak more than one language, you know the feeling when the mental switch in your head starts up with the rattling sound of a struggling engine, mixing every word you have to produce into some Denglish, Frenglish, or Spanglish. Work-related thoughts come back from your inner search engine of a brain in English, life wisdom - in the mother tongue, and the mix is unpredictable, a little weird, but it works. ## Qdrant Releases Large-Scale, Reproducible Datasets and Tooling for Vector Search Benchmarking DevFeed: [Qdrant Releases Large-Scale, Reproducible Datasets and Tooling for Vector Search Benchmarking]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-09-01T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [benchmarking](), [datasets](), [Ground truth / benchmark quality](), [Embeddings](), [reproducibility](), [Tooling](), [GPU]() Tags: [approximate-nearest-neighbor-search](), [benchmarks](), [bert](), [distributed](), [embeddings](), [fasttext](), [gpu](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [production](), [qdrant](), [recommender-system](), [reproducibility](), [research](), [saas](), [simaes-networks](), [similarity](), [tooling](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant describes Qdrant-FineWeb-10B, a large vector-search benchmarking dataset with exact top-1000 ground truth for 120,000 queries across a 10-billion-document corpus. The article also announces additional dense, sparse, and multimodal datasets and open-sources Supernova, a distributed framework for large-scale dataset generation and ground-truth computation. ### Source excerpt Real world vector search workloads are increasingly large and complex. Enterprises are not using vector search to occasionally search through a couple of PDF files. They are indexing and searching billions of vectors at thousands of requests per second (RPS) and sub 50ms tail latency. Large enterprises also can't tolerate faulty assumptions. Too many benchmarks use gated, proprietary managed services. And even worse, the data is synthetic, the queries are hidden, and the engines are locked behind paywalls. ## How to Tune Vector Search Without Guessing DevFeed: [How to Tune Vector Search Without Guessing]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-24T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [AI search](), [Qdrant](), [data]() Tags: [approximate-nearest-neighbor-search](), [bert](), [bm25](), [embeddings](), [fasttext](), [hnsw](), [how-to](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [queries](), [rank](), [recommender-system](), [retrieval](), [saas](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This article explains how to tune vector search by measuring collection settings and separating retrieval failures from ranking failures. It highlights issues involving sparse-vector weighting, BM25 configuration, candidate depth, fusion, and reranking, based on tests across five public datasets. ### Source excerpt Your collection works. Queries return in a few milliseconds, results are mostly right, and product keeps forwarding you the ones that aren't. You open the search API reference and get exact definitions for hnsw_ef, reciprocal rank fusion k, and quantization oversampling. The definitions are correct. They still don't tell you which setting is failing on your data. So you change one setting, rerun the queries, and the score moves by 0.01. Did relevance improve, or did the same queries land differently? ## How Bayer Built an Enterprise-Scale Search Engine with Qdrant DevFeed: [How Bayer Built an Enterprise-Scale Search Engine with Qdrant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-13T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI Platform](), [Artificial Intelligence](), [AI search](), [Generative AI](), [Retrieval Augmented Generation (RAG)](), [Large Language Model](), [Redis]() Tags: [ai](), [ai-platform](), [approximate-nearest-neighbor-search](), [bert](), [embeddings](), [fasttext](), [generative-ai](), [hnsw](), [image-search](), [infrastructure](), [knn-algorithm](), [large-language-models](), [large-language-models-llms](), [matching](), [neural-network](), [production](), [qdrant](), [recommender-system](), [saas](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview A case study of Bayer's enterprise generative AI platform, myGenAssist, and its use of Qdrant for vector search. The article describes the platform's growth from an initial drug-discovery application to company-wide use, with retrieval designed to support large language models, handle sustained workloads, and reduce hallucinations. ### Source excerpt Bayer is a global life sciences company operating at the intersection of two of the most consequential fields in human life: health and nutrition. Its pharmaceutical work supports drug discovery and patient care, while its crop science work supports food production at planetary scale. The company's guiding ambition, "Health for all, hunger for none," frames how it thinks about technology: AI is not a side project, but a lever applied across the entire organization, from improving the productivity of colleagues to accelerating yield prediction and drug discovery. ## Qdrant and Minima Deliver 2.92x More Agentic RAG Tasks per GPU-Hour DevFeed: [Qdrant and Minima Deliver 2.92x More Agentic RAG Tasks per GPU-Hour]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-13T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Retrieval Augmented Generation (RAG)](), [Qdrant](), [Latency](), [Inference](), [hybrid-search](), [GPU](), [Blackwell GPU](), [payload]() Tags: [approximate-nearest-neighbor-search](), [bert](), [blackwell-gpu](), [embeddings](), [fasttext](), [gpu](), [hnsw](), [hybrid](), [image-search](), [inference](), [knn-algorithm](), [latency](), [matching](), [neural-network](), [payload](), [qdrant](), [rag](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This case study describes how Minima built a bounded retrieval agent with Qdrant hybrid search, payload filtering, late-interaction reranking, and Qwen3.6-27B generation. Across evaluated tasks and agent episodes, the stack improved throughput, latency, and grounded task success compared with dense retrieval using BF16 inference. ### Source excerpt Reducing Retrieval and Calls When a retrieval-augmented generation (RAG) agent runs, it often has to plan a search, check the evidence it gets back, and try again when that evidence falls short. Those inefficiencies compound. Every extra retrieval and every extra model call adds latency, context, and inference cost. ## How to Clean Up a Qdrant Collection DevFeed: [How to Clean Up a Qdrant Collection]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-10T00:00:00Z Content type: tutorial Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [data]() Tags: [agent](), [approximate-nearest-neighbor-search](), [bert](), [data](), [embedding](), [embeddings](), [fasttext](), [hnsw](), [how-to](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [pipeline](), [points](), [qdrant](), [query](), [recommender-system](), [results](), [retry](), [saas](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview A tutorial on cleaning Qdrant vector collections whose repeated ingestion creates duplicate or stale points that displace useful search results. It explains how to inspect collection contents, deduplicate records, and use stable point IDs to make retries idempotent. ### Source excerpt Every crawl, retried job, and embedding pipeline change writes points into a vector collection. The stored data keeps moving even when the query code never changes, and the top results move with it. At first, little looks wrong. Search returns results, and latency stays normal. In our baseline run, the context-relevance score sat at 0.92 out of 1.00 while four answers in ten came back wrong, because duplicate chunks and one outdated record were filling the five results the agent could read. ## Pre-Filtering vs Post-Filtering (and Why Qdrant Does Neither) DevFeed: [Pre-Filtering vs Post-Filtering (and Why Qdrant Does Neither)]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-07T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [Benchmark](), [Graphs]() Tags: [approximate-nearest-neighbor-search](), [benchmark](), [bert](), [embeddings](), [fasttext](), [filter](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This article compares pre-filtering and post-filtering for vector search and explains Qdrant's in-place filtering approach. It describes how metadata filters can reduce recall and how Qdrant uses filtered HNSW traversal and graph repairs to preserve search quality. ### Source excerpt Adding a metadata filter to vector search can make good results disappear without making the query look broken. It still runs fast, returns something, and keeps the dashboards quiet, while some of the true nearest matches drop out. In the benchmark behind this post, a broad-value filter lowers recall to 90.8% and an AND filter over two broad values lowers it to 39.7%, while every other filter shape stays above 97%. ## Qdrant 1.19 - TurboQuant Datatype & Memory Tiers DevFeed: [Qdrant 1.19 - TurboQuant Datatype & Memory Tiers]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-08-05T01:00:00Z Content type: release Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [Compression](), [quantization](), [cloud-infrastructure](), [Collections]() Tags: [approximate-nearest-neighbor-search](), [bert](), [bm25](), [compression](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [precision](), [qdrant](), [quantization](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [storage](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant 1.19 introduces the Turbo4 datatype, which stores vectors as 4-bit TurboQuant representations without retaining full-precision copies, reducing storage by up to nine times. The release also adds unified memory tiers, tenant-specific IDF statistics, filtering improvements, and Web UI enhancements. ### Source excerpt Qdrant 1.19.0 is out! Let's look at the main features for this version: TurboQuant Datatype: A new storage format that compresses vectors to four bits without keeping their original full-precision representation, reducing storage by up to nine times compared to TurboQuant quantization. Memory Tiers: A single memory parameter unifies per-component memory tier placement, with three tiers: pinned, cached, and cold. Per-Tenant IDF Statistics: Narrow the IDF corpus to a specific tenant so term rarity reflects that tenant's vocabulary rather than the whole dataset, improving BM25 scoring in multi-tenant deployments. ## Lessons From Building E-Commerce Search on Qdrant DevFeed: [Lessons From Building E-Commerce Search on Qdrant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-07-24T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [API](), [GitHub](), [personalization](), [recommendations](), [Cloud]() Tags: [api](), [approximate-nearest-neighbor-search](), [bert](), [bm25](), [cloud](), [e-commerce](), [embeddings](), [fasttext](), [github](), [hnsw](), [hybrid](), [image-search](), [knn-algorithm](), [latency](), [matching](), [neural-network](), [personalization](), [qdrant](), [recommendations](), [recommender-system](), [saas](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This article explains lessons from building an e-commerce search storefront on Qdrant. It describes combining dense-vector and BM25 retrieval with reciprocal rank fusion, applying filters during search, and indexing payload fields for reliable production queries. ### Source excerpt Relevance, filtering, personalization, merchandising, and recommendations usually arrive as five separate services, and the final ranking gets stitched across all of them. Each service ranks by its own rules, and none of them owns the order a shopper ends up seeing. We built Qdrant Shopping, a storefront over 5.8 million real Amazon fashion products, to find out how many of those pieces collapse into one. Every text search is a single request to Qdrant's Query API that returns a ranked shelf in about 40 milliseconds, and the code is on GitHub. The decisions below apply to almost any product catalog, and we got several of them wrong before we got them right. ## Qdrant Beats Elastic's DiskBBQ at 2x Throughput, Half the Latency, and 1/3 the Compute DevFeed: [Qdrant Beats Elastic's DiskBBQ at 2x Throughput, Half the Latency, and 1/3 the Compute]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-07-08T00:00:00Z Content type: comparison Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [benchmarking](), [Benchmark](), [elasticsearch](), [Latency](), [Embeddings](), [quantization](), [IO](), [cloud-infrastructure]() Tags: [approximate-nearest-neighbor-search](), [benchmark](), [bert](), [cpu](), [elastic](), [elasticsearch](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [latency](), [matching](), [metric](), [network-attached-storage](), [neural-network](), [qdrant](), [quantization](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant responds to Elastic's DiskBBQ benchmark for disk-based vector retrieval, arguing that the comparison omitted Qdrant's two-stage retrieval and asynchronous disk scoring. In its own tests at matched recall, Qdrant reports about twice the throughput, roughly half the latency, and about one-third of the per-node CPU and RAM usage. ### Source excerpt TL;DR Elastic recently published a benchmark claiming that their proprietary, disk-based index (dubbed "DiskBBQ") delivers up to 7x higher throughput than Qdrant when deployed on nodes with network-attached storage. Elastic set out to benchmark DiskBBQ against a Qdrant cluster configured for disk-based retrieval, but their methodology omitted the exact features Qdrant built for this workload. Instead of enabling our documented two-stage retrieval and async disk scoring, they effectively ran a stress test on unbounded sequential disk access and reported the resulting I/O bottleneck as a baseline metric. ## Branch-Aware Semantic Code Search with Qdrant DevFeed: [Branch-Aware Semantic Code Search with Qdrant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-07-02T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [code search](), [Qdrant](), [Git](), [AI Agent](), [Code](), [Parser](), [Tree-sitter](), [Rust]() Tags: [agent](), [ai-agent](), [approximate-nearest-neighbor-search](), [bert](), [code-search](), [coding](), [context](), [embeddings](), [fasttext](), [git](), [hnsw](), [image-search](), [index](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [recommender-system](), [rust](), [saas](), [search](), [semantic](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This article explains how to build branch-aware semantic code search with Qdrant. It describes why a vector index can return code from the wrong Git branch and outlines structural chunking, version tracking, and synchronization with Git so coding agents receive branch-correct context. ### Source excerpt Most code search is lexical. You grep a string or jump to a definition, and because it runs on your checkout, it always reflects the branch you have open. Semantic code search goes further: you index the codebase as vectors and search by meaning, which is how you hand an AI agent the right context in one lookup instead of a long grep loop. But it brings a problem grep never has. ## Qdrant Lands in SF: Vector Space Day 2026 Recap DevFeed: [Qdrant Lands in SF: Vector Space Day 2026 Recap]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-06-24T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [hybrid-search](), [Retrieval Augmented Generation (RAG)](), [Embeddings](), [AI search](), [recommendations](), [Compression](), [GPU](), [consistency](), [Resilience]() Tags: [2026](), [approximate-nearest-neighbor-search](), [bert](), [compression](), [developers](), [embeddings](), [fasttext](), [gpu](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [memory](), [neural-network](), [product](), [product-search](), [qdrant](), [query-engine](), [rag](), [recap](), [recommendations](), [recommender-system](), [resilience](), [retrieval](), [robotics](), [saas](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant's recap of Vector Space Day 2026 in San Francisco covers discussions on agents and memory, search and retrieval, and edge and robotics. It highlights Qdrant's Rust core, storage layer, GPU-accelerated indexing, vector compression, and use cases including semantic search, RAG, recommendations, and product search. ### Source excerpt Recap Watch all the videos On June 11th, 2026, over 350 developers, researchers, and engineers came together at The Midway in San Francisco for Vector Space Day, our first event of its kind in the United States and our first major gathering in San Francisco. This was a single day, single stage, across three tracks: Agents and Memory, Search and Retrieval, and Edge and Robotics. Hosted by our MC for the day, Adam Chan, who kept the energy flowing from opening keynotes to the final hackathon reveal. ## Memory at the Edge: On-Device Vector Search with Qdrant Edge DevFeed: [Memory at the Edge: On-Device Vector Search with Qdrant Edge]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-06-16T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [Artificial Intelligence](), [Library](), [Rust]() Tags: [ai](), [approximate-nearest-neighbor-search](), [bert](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [on-device](), [qdrant](), [recommender-system](), [retrieval](), [rust](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant Edge brings Qdrant vector search onto devices as an embedded Rust library. The article demonstrates an on-device robot memory workflow that captures camera data, converts it to vectors, and searches locally without relying on a network connection. ### Source excerpt On its first day in an unfamiliar house, a home robot has to build memory as it goes: which rooms it has covered, where it last saw the car keys, whether the kitchen looks different now than it did this morning. And it has to answer those questions itself, where it stands, because the network isn't always there, and it's too slow to wait on even when it is. That memory has a concrete shape. As the robot moves, it turns what its camera sees into vectors and writes them to a store it carries onboard. To make a decision, it queries that store for the nearest matches to what it is looking at, filtered by where or when it saw them. Capture, embed, search, decide, and the loop runs entirely on the robot, in milliseconds, with no trip to a server. The engine underneath it is Qdrant Edge: the same Qdrant vector search engine, running in-process as an embedded library instead of behind an API. ## Vector Space Hackathon 2026 DevFeed: [Vector Space Hackathon 2026]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-06-11T00:00:00Z Content type: news Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [AdventureX 2025](), [Qdrant](), [Embeddings](), [AI search](), [PyTorch](), [API]() Tags: [2026](), [api](), [approximate-nearest-neighbor-search](), [bert](), [embeddings](), [fasttext](), [hackathon](), [hnsw](), [image-search](), [infrastructure](), [knn-algorithm](), [matching](), [neural-network](), [pytorch](), [qdrant](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [technical](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant announces the winners of its 2026 "Think Outside the Bot" hackathon, which challenged participants to develop creative applications of vector search without RAG or simple chatbots. The article describes winning projects including MemoryAtlas, which combines multimodal vectors and sequence modeling for early mental-health spiral detection, and Crowd Whisperer, which simulates real-time crowd reactions to music. ### Source excerpt Wow. So many cool and creative submissions for this year's hackathon; we really had a tough time picking only 3 winners! The submissions ranged from early mental health detection to crowd-reaction simulators, tactical football search, and infrastructure stress-testing. We're excited to share the results with you. The Hackathon Qdrant's 2026 "Think Outside the Bot" hackathon pushed the creative boundaries of vector search. Participants from around the world were challenged to create innovative uses of Qdrant, without the use of RAG or simple chatbots. Submissions were judged on the criteria of Innovation, Creativity, and Technical Depth. The hackathon ran for 5 weeks with winners announced at Vector Space Day 2026 with a total of $10k in prizes. Keep reading to learn about the winning submissions. ## How Sunny Health Built an AI Healthcare Concierge with Qdrant DevFeed: [How Sunny Health Built an AI Healthcare Concierge with Qdrant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-05-21T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Artificial Intelligence](), [Qdrant](), [data](), [Structured-data](), [Single sign-on (SSO)](), [Availability]() Tags: [ai](), [approximate-nearest-neighbor-search](), [availability](), [bert](), [data](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [postgres](), [qdrant](), [recommender-system](), [retrieval](), [saas](), [scale](), [schema](), [simaes-networks](), [similarity](), [sso](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview A case study of Sunny Health's AI healthcare concierge, which uses Qdrant as a retrieval layer for benefits navigation, provider matching, and appointment booking. The article describes why the team moved beyond Postgres as provider data scaled and its schema became more complex. ### Source excerpt Most people don't read their insurance pamphlet. The benefits are there: deductibles, copays, in-network providers, what dental covers, what dermatology covers, when an optometry visit is included in the medical plan. But the document is dense, the website is worse, and the result is that patients pay for plans they barely understand and delay care because finding an in-network provider with availability takes more energy than they have. Sunny Health is building a healthcare concierge that insurance companies and care providers offer to their members as part of the existing plan experience. When a member signs in (typically through SSO from their payer), Sunny Health already knows who they are and what their plan covers. They land in a chat experience where they can ask "show me dermatologists nearby," get matched to in-network options, and have Sunny Health book the appointment on their behalf. Three things on one retrieval layer: benefits navigation, provider matching, and appointment booking. ## How GoPerfect Built an Agentic Recruiting Workforce with Qdrant Cloud DevFeed: [How GoPerfect Built an Agentic Recruiting Workforce with Qdrant Cloud]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-05-19T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [Embeddings](), [Artificial Intelligence](), [Large Language Model](), [data](), [Cloud]() Tags: [agentic](), [ai](), [approximate-nearest-neighbor-search](), [bert](), [embedding](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [llms](), [matching](), [neural-network](), [qdrant](), [recommender-system](), [retrieval](), [saas](), [semantic-search](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview This case study describes how GoPerfect built an agentic recruiting platform using Qdrant Cloud, combining large language model reasoning, semantic search, vector retrieval, embeddings, and structured filtering. The system processes a large talent corpus to produce more accurate candidate shortlists and automate recruiting workflows. ### Source excerpt GoPerfect mission is to use an AI recruiting workforce that replaces the manual, low-leverage parts of recruiting. Instead, an agent decomposes recruiter intent and runs the work end to end to find top talent. Their agentic platform handles sourcing, scanning, reviewing, outreach, admin work as well as candidate conversations for recruiters, hiring managers, agencies, and CEOs who hire at volume. Recruiting is a needle-in-a-haystack problem with two complications: the haystack is massive (200M+ profiles enriched with 1B+ data points drawn from professional networks, code repositories, company data, and AI-derived signals), and the definition of the "needle" is more nuanced than any keyword filter can express. A product manager is not a product marketer, even though the two sit close together in any reasonable embedding space. ## How Sapu Indexed 28 Million PubMed Abstracts to Accelerate Cancer Research with Qdrant DevFeed: [How Sapu Indexed 28 Million PubMed Abstracts to Accelerate Cancer Research with Qdrant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-05-12T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [AI Development](), [AI Platform](), [Self-hosted](), [Docker]() Tags: [ai](), [approximate-nearest-neighbor-search](), [bert](), [docker](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [recommender-system](), [reliability](), [saas](), [self-hosted](), [simaes-networks](), [similarity](), [tooling](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Sapu built an internal AI platform for searching, synthesizing, and querying biomedical research documents. As its corpus and workloads grew, it moved from self-hosted Qdrant on Docker to Qdrant Cloud Premium after encountering infrastructure stability and expertise constraints. ### Source excerpt Sapu is an early-stage biopharmaceutical company developing treatments for hard-to-treat cancers. From its San Diego facility, the team is pioneering a nanomedicine pipeline that takes existing FDA-approved drugs and re-engineers them at the nanoscale, making them smaller, more effective, and less toxic. Building on already-approved compounds gives Sapu a stronger and faster path to therapeutic success in an industry where most candidates never reach patients. Behind the lab work sits an AI tooling suite that does the reading, searching, and synthesis that would otherwise take researchers thousands of hours. Sapu's internal AI platform supports research paper authorship, references standard operating procedures, and lets the team query its document corpus with the precision biotech R&D requires. As the company grew, so did the volume of documents, the variety of use cases, and the demands placed on the underlying retrieval infrastructure. ## Qdrant 1.18 - TurboQuant DevFeed: [Qdrant 1.18 - TurboQuant]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-05-11T08:00:00Z Content type: release Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [quantization](), [Compression](), [Monitoring](), [API](), [Web]() Tags: [api](), [approximate-nearest-neighbor-search](), [bert](), [compression](), [embedding](), [embeddings](), [fasttext](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [monitoring](), [neural-network](), [qdrant](), [quantization](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [web](), [word2vec]() ### AI overview Qdrant 1.18.0 introduces TurboQuant, a quantization method designed to provide higher compression while maintaining comparable recall and speed. The release also adds memory monitoring through the Web UI and API, named-vector schema changes without collection recreation, audit-log querying, request tracing IDs, per-collection API metrics, and strict-mode guardrails. ### Source excerpt Qdrant 1.18.0 is out! Let's look at the main features for this version: TurboQuant: A new quantization method that, at twice the compression ratio of scalar quantization, delivers similar recall and speed. Memory Monitoring: Inspect a collection's disk, RAM, and page cache usage broken down by component (vectors, payload, indexes, and more) via a new Web UI view and API endpoint. Adding and Removing Named Vectors: Add or remove named vectors to an existing collection's schema without having to recreate it. ## Presenting Sentinel - Gen AI Zürich Hackathon Winner DevFeed: [Presenting Sentinel - Gen AI Zürich Hackathon Winner]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-04-29T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Artificial Intelligence](), [AI search](), [gen ai](), [Qdrant](), [Hackathon](), [Database](), [FastAPI](), [Python](), [d3.js](), [React](), [GitHub](), [Cloud]() Tags: [ai](), [approximate-nearest-neighbor-search](), [bert](), [d3-js](), [embeddings](), [fastapi](), [fasttext](), [frontend](), [gen-ai](), [github](), [hackathon](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [python](), [qdrant](), [real-time](), [recommender-system](), [risk](), [saas](), [simaes-networks](), [similarity](), [transformer](), [users](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview The article presents Sentinel, an AI-powered system built during the Gen AI Zürich Hackathon to detect contradictions across news sources. It describes Sentinel's use of scheduled news collection, Qdrant semantic search over article embeddings, and Mistral-7B analysis to identify factual inconsistencies and calculate misinformation risk scores. ### Source excerpt When Ali Aoun Mehdi watched two major global news outlets report opposite facts during the Iran-US conflict, he saw a critical problem: misinformation spreading in real-time. From Islamabad, participating virtually in the Gen AI Zürich Hackathon, he built Sentinel to close this "fact-gap" by detecting contradictions across news sources instantly. Sentinel is an AI-powered early warning system that monitors 20 global news sources every 30 minutes. It identifies factual contradictions in under 30 seconds, providing users with a misinformation risk score and narrative traction assessment. ## Now Available on Qdrant Cloud: GPU Indexing, Multi-AZ, and Audit Logging DevFeed: [Now Available on Qdrant Cloud: GPU Indexing, Multi-AZ, and Audit Logging]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-04-28T00:00:00Z Content type: release Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [GPU](), [Availability](), [Logging](), [Replication](), [API](), [JSON](), [SIEM, Security]() Tags: [accelerated](), [api](), [approximate-nearest-neighbor-search](), [availability](), [bert](), [embeddings](), [fasttext](), [gpu](), [hnsw](), [image-search](), [json](), [knn-algorithm](), [logging](), [matching](), [neural-network](), [qdrant](), [real-time](), [recommendation-systems](), [recommender-system](), [replication](), [saas](), [security](), [simaes-networks](), [similarity](), [systems](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant Cloud announces GPU-accelerated HNSW indexing, Multi-AZ deployments, and audit logging. GPU indexing targets high-write workloads, Multi-AZ improves availability, and audit logs record Qdrant API operations with structured JSON and support SIEM integration. ### Source excerpt Now Available on Qdrant Cloud: GPU Indexing, Multi-AZ, and Audit Logging We're excited to announce some Qdrand Cloud upgrades to address AI workloads that write continuously, must meet higher uptime SLAs, and require audit logging. Speed Indexing with GPUs GPUs aren't just for model inference; they're for indexing too. Qdrant Cloud GPU-accelerated indexing delivers up to 4x faster HNSW index builds on dedicated GPUs, based on Qdrant benchmarks. That's helpful for high-write workloads (e.g. dynamic content catalogs, real-time recommendation systems, agentic). ## How Data Graphs Built a True Hybrid Graph RAG Platform DevFeed: [How Data Graphs Built a True Hybrid Graph RAG Platform]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-04-22T00:00:00Z Content type: article Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [graph-database](), [Retrieval Augmented Generation (RAG)](), [Embeddings](), [Artificial Intelligence](), [AI Agent](), [workflow automation](), [data]() Tags: [ai](), [ai-agent](), [approximate-nearest-neighbor-search](), [bert](), [database](), [embeddings](), [fasttext](), [graph-database](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [microsoft-graph](), [neural-network](), [qdrant](), [rag](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec](), [workflow-automation]() ### AI overview A case study of Data Graphs, a knowledge-graph-as-a-service platform that combines graph databases, full-text search, vector embeddings, workflow automation, and an agentic AI layer. It argues that effective Graph RAG should combine structured graph queries with semantic retrieval instead of relying on vector search alone. ### Source excerpt Data Graphs is a UK-based platform company that provides a knowledge graph-as-a-service polystore, built on a proprietary, high-performance graph database engine. Co-founded by Paul Wilton over 10 years ago as a consultancy, Data Graphs has evolved into a platform that serves industries ranging from sports media and publishing to GLAM (galleries, libraries, archives, museums), Agri-Tech, and Regulatory Technology. The Data Graphs platform combines a proprietary graph database, full-text search, vector embeddings, workflow automation, and an Agentic AI layer into a single, unified backbone for enterprise data. For organizations managing complex, highly connected data, the platform acts as both the system of record and the context layer that reasons across it. ## Announcing Vector Space Day 2026 in San Francisco DevFeed: [Announcing Vector Space Day 2026 in San Francisco]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-04-21T00:00:00Z Content type: release Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [Retrieval Augmented Generation (RAG)](), [Hackathon](), [networking](), [Robotics]() Tags: [2026](), [agentic-ai](), [ai](), [approximate-nearest-neighbor-search](), [bert](), [developers](), [edge](), [embeddings](), [fasttext](), [hackathon](), [hnsw](), [image-search](), [knn-algorithm](), [matching](), [neural-network](), [qdrant](), [rag](), [recommendations](), [recommender-system](), [retrieval](), [robotics](), [saas](), [san-francisco](), [search](), [simaes-networks](), [similarity](), [transformer](), [vector](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant announces Vector Space Day 2026, a full-day developer event in San Francisco focused on retrieval, vector search infrastructure, agentic AI, robotics, and related systems. The document also describes a global virtual hackathon held before the event. ### Source excerpt Vector Space Day 2026: Powered by Qdrant Recap Watch all the videos Read the blog recap About We're hosting our second-ever full-day in-person Vector Space Day (https://luma.com/vsd-sf) on June 11th at The Midway in San Francisco, and you're invited. Last year in Berlin we brought together 400+ engineers, researchers, and AI builders to explore the cutting edge of retrieval, vector search infrastructure, and agentic AI. And we're excited to now bring that to San Francisco. ## Qdrant Skills for AI Agents DevFeed: [Qdrant Skills for AI Agents]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-03-31T00:00:00Z Content type: release Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [Retrieval Augmented Generation (RAG)](), [Embeddings](), [Latency](), [Artificial Intelligence]() Tags: [ai-agents](), [approximate-nearest-neighbor-search](), [bert](), [embeddings](), [engineering](), [fasttext](), [hnsw](), [image-search](), [infrastructure](), [knn-algorithm](), [latency](), [matching](), [neural-network](), [qdrant](), [rag](), [recommender-system](), [saas](), [simaes-networks](), [similarity](), [transformer](), [vector-database](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant introduces skills that help AI agents and humans make production vector-search decisions. The article explains tradeoffs involving quantization, HNSW tuning, payload filtering, hybrid retrieval, sharding, replication, storage, memory, latency, and ranking quality, with Cosmos as a visual-search example. ### Source excerpt The standard RAG tutorial teaches a simple pattern: embed your documents, store them in a vector database, retrieve the top K, and feed them to the LLM. The vector engine is passive infrastructure. Put vectors in, get neighbors out. Configure once, forget about it forever. This mental model is why most AI agents treat vector search as a black box. They can call the API. They cannot make the engineering decisions that determine whether it works well. ## Qdrant Introduces a Free Course on Production Multi-Vector Search DevFeed: [Qdrant Introduces a Free Course on Production Multi-Vector Search]() Original publisher: [Read original article]() Author: info@qdrant.tech (Andrey Vasnetsov) Published: 2026-03-24T00:00:00Z Content type: tutorial Language: en Sources: [Qdrant Blog on Qdrant - Vector Search Engine]() Topics: [Qdrant](), [AI search](), [Embeddings](), [Machine learning](), [Optimization](), [Python](), [quantization](), [API]() Tags: [advanced](), [api](), [approximate-nearest-neighbor-search](), [backend](), [bert](), [embeddings](), [evaluation](), [fasttext](), [hnsw](), [image-search](), [implementation](), [knn-algorithm](), [matching](), [neural-network](), [production](), [qdrant](), [recommender-system](), [saas](), [scale](), [search](), [simaes-networks](), [similarity](), [transformer](), [tutorials](), [vector-search](), [vector-search-engine](), [vectors](), [word2vec]() ### AI overview Qdrant introduces a free advanced course on multi-vector retrieval for ML, backend, and search engineers. The course covers late interaction models, text and multimodal search, optimization, evaluation, and production implementation with Qdrant. ### Source excerpt Most vector search tutorials stop at single-vector embeddings: one document, one vector, one similarity score. That works for demos. It falls apart when your retrieval pipeline needs to capture fine-grained token-level interactions across text, images, and PDFs at production scale. Until now, engineers who wanted to go deeper had to piece together scattered papers, blog posts, and half-documented repos. There was no structured, hands-on resource that connected the theory of late interaction models to real implementation in a production search engine. [Next page]()