The Good, the Bad, and the Leaky: jemalloc, bumpalo, and mimalloc in Meilisearch
The good, the bad, and the leaky: jemalloc, bumpalo, and mimalloc in Meilisearch
Clément Renault
Patching LMDB: how we made Meilisearch’s vector store 3× faster
We patched LMDB to support nested read transactions on uncommitted writes - eliminating full database scans and making Meilisearch's vector store 3× faster
Clément Renault
Why traditional hybrid search falls short (and how we fixed it)
Meilisearch's innovative scoring system revolutionizes hybrid search by properly combining full-text and semantic search, delivering more relevant results than traditional rank fusion methods.
Louis Dureuil
Introducing Meilisearch's next-generation indexer: 4x faster updates, 30% less storage
Indexer edition 2024 revolutionizes search performance with parallel processing, optimized RAM usage, and enhanced observability. See what's new in our latest release.
Louis Dureuil
Meilisearch indexes embeddings 7x faster with binary quantization
By implementing binary quantization with the vector store Arroy, significant reductions in disk space usage and indexing time for large embeddings have been achieved while maintaining search relevance and efficiency.
Tamo
Meilisearch is too slow
In this blog post, we explore the enhancements needed for Meilisearch's document indexer. We'll discuss the current indexing engine, its drawbacks, and new techniques to optimize performance.
Clément Renault
How Meilisearch updates a database with millions of vector embeddings in under a minute
How we implemented incremental indexing in our vector store.
Tamo
Meilisearch expands search power with Arroy's Filtered Disk ANN
How we implemented Meilisearch filtering capabilities with Arroy's Filtered Disk ANN
Clément Renault
Multithreading and Memory-Mapping: Refining ANN performance with Arroy
Overcoming the challenges to enhance ANN performance with Rust.
Clément Renault
Spotify-Inspired: Elevating Meilisearch with Hybrid Search and Rust
How we created Arroy, a Rust library building upon the foundations of Spotify's Annoy.
Clément Renault
From ranking to scoring
Our journey to add relevancy scores to search results in Meilisearch.
Carolina Ferreira
Squeezing millions of documents in 128 TB of virtual memory
How dynamic management of virtual memory enabled us to remove limitations in Meilisearch indexing policy.
Louis Dureuil
Prose linting with Vale
Learn the importance of a style guide in technical writing and how Vale ensures your content complies with it.
Maryam Sulemani
Automate pull requests merging with Bors
How we started using Bors to create a better environment for Meilisearch's contributors.
Clémentine Urquizar
SpO₂ the little dynamic monitoring tool
Introducing SpO2, an open-source dynamic monitoring tool built in Rust for health-checking Kubernetes pods with real-time Slack notifications.
Clément Renault
MeiliES - Event sourcing in Rust
Introducing MeiliES, an event sourcing database built in Rust using the Redis protocol, created as a lightweight alternative to EventStore.
Thomas Payet