Datasets containing input thermodynamic properties and parameters are needed in models used to predict the chemical evolution of natural systems, material properties, and industrial processes. New and updated data are constantly being produced, and many distributed datasets are only available as part of a modelling code package in different data formats, making them difficult to track, compare and apply. The project aimed to enhance and expand the ThermoHub (https://thermohub.org/thermohub) database and associated tools to ensure to ensure the traceability and completeness of thermodynamic datasets. This is done by creating a unified thermodynamic database in a general format, following the FAIR principles, and developing a workflow for curating and uploading new data and datasets, providing a path for further continuous extensions and maintenance. The work focused on improving the open source tools for data processing, writing and adapting the import scripts for each dataset with its specific data format and adapting the schemas that describe the data types and make them machine readable. A semiautomatic workflow was developed and adapted to import data from several mainstream databases used in chemical modeling and link them to their corresponding bibliographic source, taking advantage of the graph structures in ThermoHub. Datasets can be kept up to date and the workflow can be run again when data are updated or it can be used as template and adapted for other datasets and formats. Newly imported data can be used by modeling codes like GEMS (developed at PSI https://gems.web.psi.ch/) and Reaktoro (developed at ETHZ https://reaktoro.org/) extending their applicability to various chemical systems and over wide ranges of conditions.