The FalconV engine combines machine and deep learning techniques throughout the funnel. Various neural network types, tailored ensembles and techniques like transfer learning (e.g. BERT and FastText) combined with other approaches like statistical analysis and complex algorithms are utilized to facilitate a variety of microservices.

AI setup

NOW
  • Assistant 2.0
  • Automated Data Extraction from semi structured documents
  • Blind Vision
  • Reliable layout perception based on a hybrid graph model approach
NOW
  • Category Moderation
  • Duplicate Detection
  • Related Content Detection
  • Style Transferred Photos
SOON
  • Cross-correlation Detection
  • Fake News Detection
  • Anomaly Detection
  • Real-time consensus modelling
  • Automatic Content-based Poll Generation
  • Text Summarization
  • AI-generated content assistant
  • Sensitive Information Discovery
NOW
  • Category Moderation
  • Duplicate Detection
  • Related Content Detection
  • Style Transferred Photos
SOON
  • Cross-correlation detection
  • Fake News Detection
  • Anomaly Detection
  • Real-time consensus modelling
  • Automatic Content-based Poll Generation
  • Text Summarization
  • AI-generated content assistant
SOON
  • Behaviour and preference-based personalized news delivery

  • Bots for RPA and various messengers (e.g. Telegram, WhatsApp)

  • Cross-domain recommendations (e.g. targeted ads, e-commerce)

  • Data Annotation / Labelling

DOWNLOAD OUR SCIENCE PAPER

HIPSTO has been one of the early adopters of the BERT model. Recently our scientific team released a paper on how to use and improve BERT for some specific task on the topic of semantic similarities of news content. please download the paper here.

Contact
close slider

HIPSTO Operating B.V.

Email: info@hipsto.global

Phone: +31 (0)6 507 458 65


Bank
IBAN: NL36INGB0007989456
BIC:INGBNL2A