Bottom-up synthetic biology aims to engineer artificial systems exhibiting biomimetic struc- ture and functionality from the rational combination of molecular and nanoscale elements. These systems often take the form of artificial cells: micron-scale, cell-like constructs that display advanced behaviours such as communication, synthesis, storage of molecules, adap- tation and motion. Artificial cells require micro-compartmentalised architectures to host and maintain separation between the various signal-processing and functional elements underpin- ning their responses. Compartmentalisation can be achieved with different solutions, most of which rely on semi-permeable membranes. However, membrane-less implementations based on hydrogels, biomolecular condensates, or coacervates are gaining traction due to their advantages in terms of design versatility and robustness. DNA nanotechnology applies our understanding of the properties of nucleic acids to program assembly and disassembly of molecular and nanoscale complexes. This has led to the creation of intricate constructs with programmable structure and kinetic response. These advantages, along with sensitivity to environmental stimuli, aptameric targeting, and ease of chemical functionalisation, have made DNA nanosystems widely used in bottom-up synthetic biology. Among the many classes of DNA-based building blocks, branched DNA motifs have been shown to aggregate into hydrogels and crystal phases. Among possible implementations, cholesterol-modified DNA junctions have been shown to robustly self-assemble into cell-size condensates with programmable features, including pores size, chemical functionality and integration of responsive molecular circuitry. These properties make amphiphilic DNA condensates promising as scaffolds for artificial cells. In this thesis I introduce a platform for engineering artificial cells reliant on self-assembled amphiphilic DNA condensates, specifically tackling two key challenges: the establishment of internal compartmentalisation and the design of inter-cell communication pathways. To engineer internal architecture, I rely on controllable reaction-diffusion processes. I show that these can generate chemically addressable domains within the condensates, with control- lable number, shape and molecular makeup. The patterning processes can be rationalised and guided by numerical modelling. As a proof-of-concept, I apply this technique to construct a prototypical artificial cell displaying spatial separation of functionality, namely, a nucleus capable of synthesising RNA and a cytoplasm-like storage domain which can accumulate it. As a means of establishing self-sustaining long-range communication between two pop- ulations of condensates, I took steps towards the design of molecular circuits capable of non-enzymatic signal amplification. Overall, my work showcases the potential of DNA condensates as a platform for bottom-up synthetic biology, which can unlock the design of biomimetic systems with ever increasingly advanced functionalities.