N-Waves

We apply Deep Learning to Natural Language Processing and intersection of NLP with Computer Vision (like text generation from Images). Our research focuses on adoption of break-throughs in English NLP to local markets and custom data sets. Recently, we've adopted ULMFiT, an architecture that has state of the art in document classification of English, to Polish and German. It brought us the first prize in PolEval2018, language modeling task. We've achieved new state of the art in Sentiment Analysis on GermEval2017 and found a unified solution that gets state of the art results on both GermEval2018 tasks (hate speech detection). Practical, business applications of NLP we find the most enjoyable and we’d love to see it's working for you. Recently the question “if AI can support my company” is better replaced with “How AI can support my company?” We offer custom deep learning models performing: Sentiment Analysis (Emotions classifier, Tweet Sentiment, Product Sentiment including specific kind of sentiments like Hotel, Airline or Restaurants), Text classification (keywords extractor, text summarising and classifying, detecting hate speech) and others, in any language (some of the languages are still under development though).