Weaviate is a smart search engine with Machine Learning capabilities. It is an API-driven data platform that can automatically classify text, makes it easier to retrieve information from data and can provide recommendations by presenting knowledge in context. For example, detecting sms-phishing, significantly better search results in your Sharepoint, offering related products in your webshop and automated classification of submitted declarations.


How does Weaviate work?

In almost every situation when it comes to data, that data refers to something in the real world. Such as a banking transaction, a vehicle, a medical indication, geographic location, etcetera. A challenge for most of the databases is the difficulty to truly understand the context of a situation. An example is the characters ‘Apple’, it can refer to a piece of fruit or to the brand.


Besides that, almost all data is related to something, as Amsterdam is the capital of the Netherlands. Weaviate does not only store the concept but also its relation to other concepts; the city of Amsterdam is related to the country the Netherlands. This means that the data which is added to Weaviate, creates a network of knowledge and context.


The next big step in search happens when Machine Learning is added to traditional search methodologies. Developers can not only search for focus key words but also for context. A category which does not yet exist, can be created in process. An example is showing pumpkin products for Halloween.

Practical examples


ERP and Supply Chain systems

Data classification in ERP or Supply Chain systems is often done manually or by external data science teams. Weaviate uses Artificial Intelligence solutions to classify data and avoid expensive teams or manual labour to validate and check the data.


Document Search and Analysis

Companies mostly use unstructured documents to store text such as PDF and Word documents, website input fields and emails. The challenge today is that data can only be retrieved when searching for the right keywords rather than the context in which the data is presented. For example, when searching for “science fiction” in a traditional book search engine, you might find anything related to “science” or “fiction” so this would include “neuroscience” but not “a book about the future”.



A search bar is one of the most important functionalities of an e-commerce website. If a visitor can not find what s/he is looking for, it leads to a potential lost sale. Weaviate makes it possible to search for the context of a product as well and not only for the keyword of the product.


Cyber Security

SIEM platforms process large amounts of security related streaming data. Indexing and analysing free text is still a challenge to do on large scale causing threat mitigation happening too late or – worse – not at all. Weaviate can classify automatically if and which actions should be taken based on free text.



Semantic search

Weaviate understands both the meaning and interpretation of a text. In a banking environment you want to be able to search spending in restaurants or supermarkets. With the current search function that is not possible because you can only search by specific name. Weaviate is capable of searching for context and understands that spending in a supermarket and at the bakery are both “groceries”.


Automatic classification

Weaviate links things together automatically without any manual action. For example, an iPhone belongs to the category “Telephone” but also to the category “Apple products”.


Knowledge representation

Weaviate places a characteristic or word in relation to its context. Weaviate understands that Amsterdam is the capital of the Netherlands and will place this closely together in its database model. Knowledge is captured without pre definition, Weaviate uses machine learning to do that.


Do you want more technical information about Weaviate, check the Weaviate site.

Curious about the possibilities and the benefits for your organization, please contact our experts!