I am working in a few pet projects related with AI / Deep learning and distributed technologies.
Most of them are being developed. Sorry, if they are too desorganized yet, the idea is to clean them and add a good test coverage when they start to be mature and stable.
These projects have been developed with a complete stack:
- VueJS - frontend
- Flask - Backend
- Gunicorn and Docker for deployment
- Keras, Tensorflow or Pytorch for deeplearning
- SpaCY or Gensim for NLP The most interesting ideas which I am working are:
A very simple service for CBIR - Retrieval systems. Applying VGG16 networks, we calculate embedding to find similar images.
Applying RNN networks and websockets, I created a chatbot to talk (trained model from tweets)
Designed web crawler which analyse different news to detect useful information (topics, sentiment, etc)
My motivations in JAVA is using different architectures. The basic example is the implementation of a clean architecture: link