Introducing Kvasira

Published 26 Aug 2019 by Kvasir Analytics

KvasirA: meaning-full search, instantly 🚀

It is easy to become overloaded from the many high volume information flows that exist on the Internet. We are excited to announce KvasirA, a smart content discovery tool to help you manage this rising information flood.

Traditionally, searching for related content to the document you’re reading requires you to read enough to figure out some relevant keywords, and to enter those keywords into a search engine. For all but the simplest documents, you’ll probably need to iterate this process several times, unless you’re already intimately familiar with the knowledge domain:

KvasirA enables a different, more efficient, workflow that uses the document you are reading as the query directly. KvasirA takes the document you provide, matches it against documents in one or more of the libraries it maintains, and returns the semantically closest documents – that is, those with a similar meaning:

At the heart of KvasirA are a set of document libraries. Each library is created by ingesting documents in a variety of common formats and can be incrementally updated over time. Each document passes through a machine learning pipeline that is completely automated and managed by us — no need to train your own models or tweak parameters!

To query a document library, you simply provide a document of your own. KvasirA supports a range of formats including plain text, PDF, or a link to a webpage. KvasirA then returns documents similar to the query from the chosen document library. For example, we can query Wikipedia with the URL

KvasirA provides a straightforward interface for creating and querying document libraries via our website, along with a growing number of integrations into third-party tools provided using our flexible API. Our API enables KvasirA to be used for a variety of applications, such as:

  • Document search and management: Index your documents with KvasirA and use it to present a system for providing a fast search function with minimal setup.
  • Recommender systems: Embed an automated KvasirA search on a page to recommend similar content to users. For example, KvasirA can be used to present the user similar news articles based on the one they are currently reading.
  • Exploratory search: When you don’t know what you’re looking for, you can perform iterative searches using the engine to discover new and interesting content from a given starting document.

KvasirA makes it easy to create and curate pipelines for your own document libraries in over 10 languages, including English, French, Spanish, and Arabic. Your query document does not even need to be in the same language as the document library you’re matching against!

Give KvasirA a try using one of our curated document libraries at Coming soon: our Slack integration!

We’re working hard to release new integrations and features. If you’re interested in a full demo, indexing your own documents or think you have a great use for KvasirA, drop us an email at