Social Thermometer

End-to-end development of WebApp for the analysis of tweets published in Uruguay in real-time.

  • Natural Language Processing (NLP)

    We train the BERT model in Spanish to perform Sentiment Analysis of each Tweet. Model put into production with Google AI Platform (GCP).

  • Community Detection

    Implementation and development of Machine Learning algorithms linked to graph theory for detection of user communities and production interactively with JavaScript and WebGL.

  • Automated Pipeline on DataFlow.

    All data flow is automated from GCP's DataFlow (Apache Beam) and Pub / Sub.

  • The infrastructure developed entirely in GCP.

    All the infraestructure was developed integrating the GCP service stack (DataFlow, Pub/Sub, Vertex AI, BigQuery, CloudBuild).