Relational databases and the associated SQL language are rapidly being supplanted in modern data systems by so-called "NoSQL" databases, most commonly key-value stores, document databases, and graph databases. These models often provide greater flexibility, simplicity, and efficiency when representing various sorts of structural data. Combined with the benefits of Rust, data can be represented more naturally and processed far faster using NoSQL.
Traditional statistical machine learning and modern deep learning methods are essential parts of many modern back-end systems nowadays, but getting top performance out of them is often a challenge. We can take your data-scientific findings or prototypes and implement them in highly-efficient and scalable Rust code, in addition to building ML/AI systems from scratch.