Deep learning is a subfield of machine learning that uses deep neural networks – models with many layers – to detect complex patterns in large datasets. It is the technology behind modern language models, image processing, speech and text recognition, generative AI and much more.
Deep learning differs from classic ML mainly in two ways: the depth and size of the models, and the ability to learn representations from raw data instead of constructing them by hand. This has rendered manual feature engineering partly obsolete and massively expanded application fields.
In recruiting, resume parsers, semantic search, voice and image analysis benefit from deep learning. At the same time the demands on training data, compute and explainability grow. Deep models are often hard to interpret, which raises bias risks and tightens regulatory duties in the EU.
Lunigi uses deep-learning-based models in semantic search, without requiring candidates to handle the technology themselves.