In recent years, data-driven methods have started to play an important role in the field of astrophysics. In the context of large Milky Way spectroscopic surveys such as GALAH, APOGEE, RAVE, and soon Gaia-RVS, WEAVE, and 4MOST, these tools are key in parameterizing millions of spectra in a short time.

In this talk, I will show that a Convolutional-Neural-Network-based approach (CNN) offers a unique way of combining spectroscopic, photometric and astrometric data smoothly. CNNs are known to be key for detecting patterns and features in (two dimensional) images, for example. In the present study, we work with one dimensional stellar spectra, characterized by spectral line features. Such features are indicators of the physical properties of the stars (temperature, gravity, chemical composition, etc). I will present a study, based on APOGEE DR16 (R=22,500) and RAVE DR6 spectra set (R=7500), the goal being to derive precise atmospheric parameters and chemical abundances for more than 400,000 RAVE spectra.