Microorganisms are ubiquitous and thrive in a multitude of environments. Organisms adapt to sustain life in these dynamic environments. They acquire new traits, thus forming the basis of evolution. Metabolic networks are useful candidates to study evolution, as they form a bridge between metabolic genotype and phenotype. In our work, we aim to uncover some of the evolutionary design principles underlying metabolism. To start with, we evolve metabolic networks in silico on glucose and study the evolution of functional redundancy on different environments. We could identify a wide variety of compensation mechanisms that exists in these evolved metabolic networks. We could also find that the starting metabolic network possess higher redundancies than that exists in these randomly evolved networks. Are these multiple redundancies by virtue of the nutrient condition, i.e. glucose? To address this question, we expanded the above study to a wide range of nutrient conditions, consisting of 9 other carbon sources. We found that evolution confers growth advantages on select nutrient conditions and also a myriad of redundancies. Further, owing to such redundancies, it is possible to minimize a metabolic network into smaller functional networks. We devise a systematic algorithm for minimizing metabolic networks considering the network structure and redundancy. We show that our approach is faster and efficient than other methods available till date. Going beyond individual organisms, we next turn towards studying the underlying principles of microbial interaction. We co-evolve several communities that exhibit different interaction types, and study them to understand the stability of the interaction types. Overall, our study throws light into some of the design principles in metabolism ranging from individual organisms to microbial communities.