Organic solvents are ubiquitous in industrial and domestic applications from the production of pharmaceuticals to household consumer products. The negative impact of most traditional solvents, especially aprotic types, on the environment, health, and safety has long been a matter of public and professional concern. These issues led to a high demand for developing solvent alternatives, preferably from renewable sources. In this context, lignocellulosic biomass is a promising feedstock since it is abundantly available and does not compete with food crops. Most approaches for producing bio-based aprotic solvents revolve around biomass-derived carbohydrates but they often suffer from low yields and/or high costs due to multi-step processes and expensive catalysts required to convert natural sugars into aprotic molecules. To address this, our research group developed the Aldehyde-Assisted Fractionation (AAF) technology that transforms carbohydrates into acetal aprotic form in a single step by protection chemistry during biomass depolymerisation. This approach originally utilised low-cost bulk chemical formaldehyde, resulting in the formation of a new platform chemical diformylxylose (DFX) that could be an interesting aprotic solvent candidate. In the first part of this doctoral thesis, I show that DFX has unique properties that enable its use as a polar aprotic solvent. I demonstrate the notable performance of DFX in organic synthesis and in other applications such as biomass pretreatment and liquid-phase exfoliation of nanomaterials. I further perform mutagenicity and toxicology assessment of DFX to confirm its potential to be a safer alternative to toxic traditional polar aprotic solvents. In the second part, I demonstrate how waste biomass can be sustainably transformed into DFX with low cost and high efficacy. Specifically, I provide a scalable method for the one-pot production of DFX from corn cobs, while adhering to the principles of green chemistry. The developed method offers an outstanding environmental profile and economic competitiveness as was proved by preliminary life-cycle assessment and techno-economic analysis. I also explore the biodegradability of DFX to evaluate its potential as an eco-friendly solvent. Inspired by the revealed potential of DFX as a solvent, I expand the portfolio of xylose-based solvents with acetal functionality in the next part. Specifically, I incorporate other short-chain aliphatic aldehydes into the xylose core to produce new xylose acetals with diverse properties from biomass. I illustrate the potential of this new class of bio-based solvents in synthetic chemistry with a particular focus on polycondensation biocatalytic reactions. In the last part of the thesis, I address the general challenges in solvent design and selection and contribute to the development of novel computational tools. The first proposed tool is a quantum mechanical framework that facilitates rational solvent selection for biorefinery processes. The second framework combines machine learning with a genetic algorithm inspired by the process of natural evolution to design new effective solvent candidates for lignin. In summary, this thesis shows the development of a new class of carbohydrate-based solvents with acetal functionality that can be sustainably produced from biomass. The described findings contribute to the overall goal of developing environmentally benign, safe, and sustainable bio-based solvent candidates.