Hi, we're NLP Town. We apply

Natural Language Processing

to help you get more useful information
from collections of unstructured text.

Sentiment analysis

Texts contain
a wealth of information

Analyzing information in texts requires a lot of effort. People have to find all relevant documents, read them and synthesize the information they contain.


Make your texts
work for you

At NLP Town, we help organizations put their texts to work. We make software that analyzes documents automatically and unlocks critical knowledge in the blink of an eye.

How can we
help you?

We take on two kinds of jobs: consulting projects and development projects. Just pick the format that works best for you.


We help you define the NLP strategy that best suits your company. We teach your people how to develop top NLP models, either in a multi-day workshop or a longer consulting trajectory.

Software & model development

Focus on what you do best and let us take on the NLP work. We develop your NLP software and train the best models for your domain. We integrate our software with your workflow.

Example Projects

Improving citizen participation

Start-up Citizenlab builds a software platform for civic engagement. They asked us to develop NLP models to analyze the suggestions citizens make to their city council. Our models classify these suggestions by topic, cluster similar suggestions together and automatically identify the places people mention.

Image Description

Analyzing sentiment for RTBF

For their popular television show A votre avis, Belgian broadcaster RTBF relies on our NLP software to search and analyze millions of tweets.

  • Semantic search
  • Sentiment analysis
Image Description

Parsing founding deeds

Together with the Federation of Notaries (FEDNOT), we developed a document parser that automatically identifies the most relevant information in founding deeds, such as the name of a company, its address and the names of its founders.

Natural language processing (NLP) is rapidly becoming as integral to the workplace as communication itself.

Image Description

Mining search logs

To improve the quality of their search engines, we helped Wolters Kluwer analyze tens of thousands of user queries, and use the findings to transform the search experience across their online products.

Blog Posts

Distilling BERT Models with spaCy

We successfully applied model distillation to train spaCy's text classifier to perform almost as well as BERT on sentiment analysis of product reviews.

  • By: Yves
  • /
  • Posted: August 26, 2019

Sentence Similarity

Word embeddings have become widespread in Natural Language Processing. They allow us to easily compute the semantic similarity between two words, or to find t...

  • By: Yves
  • /
  • Posted: May 2, 2018

NER and the Road to Deep Learning

Not so very long ago, Natural Language Processing looked very different. In sequence labelling tasks such as Named Entity Recognition, Conditional Random Fie...

  • By: Yves
  • /
  • Posted: Sep 12, 2018

Meet our founding team

Image description

Yves Peirsman

NLP Expert
Image description

Johan Leys

Software Architect