Tailoring particle interaction among individual building blocks remains an important challenge in a bottom-up assembly scheme. Elastic interactions in anisotropic fluids can be harnessed for this goal. For my thesis, I am particularly interested in investigating the role of confined liquid crystals as a dispersing medium in driving particle assembly. Nematic liquid crystals (NLCs) are made of rod-like molecules that tend to co-align with their neighbors, in a field called the director field. Anchoring refers to the orientation of the NLC molecules at the boundary. Elasticity arises where the molecules deviate from the director field, in three modes of distortion, splay, twist and bend. When a continuous director field cannot be present everywhere, topological defects form, resulting in small melted regions where the order parameter is undefined. It is well-known that particle with perpendicular anchoring generates associated defects, in the form of a Saturn ring or a dipole. These particles have unique symmetry, analogous to electrostatics. A wall with homeotropic anchoring repels a colloid with the same anchoring; yet by changing the surface topography from planar to concave, one can turn repulsion into attraction. This study is inspired by biology, in the so-called “lock-and-key” interaction. I demonstrate the ability to design precise docking sites, near an undulated boundary with peaks and valleys, for both Saturn rings and dipoles. The domain is engineered to be defect-free in order to avoid strong trapping sites. By tailoring wall curvature, I define sites of attraction and equilibrium loci for colloids that vary from near contact to several particle radii from the boundary. Particles dock in wells of similar radii obeying simple geometric argument that allows particle to maximize splay and bend matching. Wells of large radius stabilized colloids with a distorted Saturn ring. In certain cases, Saturn rings transform to dipolar configurations driven by wall interactions. I can also define sites of repulsion to propel colloids away from these boundaries and find unstable loci from which colloids depart along multiple paths. Small perturbations of colloid position allow selection among these paths. Colloids with different defects interact distinctly with these boundaries, depending on their near field director field. Finally, I demonstrate the ability of a colloid in motion, like “Goldilocks”, to select from wells of different sizes for preferred docking. Landau de Gennes (LdG) simulations, the standard numerical method in solving for the director field without prior knowledge of the position of the defect, are useful tools in elucidating our experimental findings. We have expanded the work to simulate dipoles, as well as mapping the energy landscape to calculate force field to corroborate experimental trajectories. These docking sites are useful tools in building structures. I have observed “eyelashes”, topological dipole chains that follow the local, curved director field. These beget wires that connect the groove corners to topographical features on the cell lid to yield oriented, curvilinear colloidal wires spanning the cell, following the curvature of the director field. As the groove aspect ratio changes, I find different ground states, including the ones that contain defect lines which compete with the corner. Anisotropic particles are natural extension to the spherical particles. I have also shown that ellipsoids have distinct energy landscape that depends on both their aspect ratio and orientation. The interaction relies on near-field director field matching rather than strength of particle-sourced distortion, thus the platform has the potential to be scaled down for nanoscale manipulation. In summary, I study how to guide the formation of reconfigurable structures in NLCs. This was achieved by using boundaries to mold the director field. Docking sites can be exploited for structure formation, such as wiring along the director field. Directing particles toward or away from boundaries provides new tools to steer colloid motion. The abilities to transform defect configuration allows for nano-manufacturing in the defect site.