In order to track single superparamagnetic microbeads serving as markers in biomolecular assays, high-sensitivity, sub-micron magnetic sensors which can be integrated onto Lab-on-a-chip platforms are needed. This thesis studies the realization and characterization of such magnetic sensors using Focused Electron Beam Induced Deposition (FEBID) of Cobalt in a High Vacuum (HV) Scanning Electron Microscope (SEM). In FEBID, a finely focused electron beam is used to selectively irradiate a surface where functional precursor molecules are adsorbed. Electron-impact dissociation of the adsorbates lead to non-volatile fragments being deposited at the point of irradiation, while volatile fragments desorb and are pumped away. We have investigated FEBID of Dicobalt octacarbonyl [Co2(CO)8]. The deposit consists of a nanocomposite CoC material, where Co nanocrystals 2-3 nm in size are embedded in a carbonaceous matrix. This material exhibits a scattering enhanced Extraordinary Hall Effect (EHE) which allows for high magnetic sensitivities. Combined with the inherent nanometric resolution of the deposition process, this makes this nanocomposite CoC material the material of choice for the deposition of small (sub-micron), high-sensitivity magnetic sensors. However, the magnetic field resolution is found to be highly dependent on the exact composition of the deposition process. We report for the first time the controlled tuning of the composition of deposits from Co2(CO)8 using the electron beam pulse time as the process parameter. We explain the tunability in terms of co-deposition of chamber background hydrocarbons. A novel, general model describing the electron-induced deposition in terms of surface adsorbate densities in the presence of two adsorbate species is presented. The analytical model allows to describe the conditions for a broad tunable composition window for any two-adsorbate system. The model is applied to the two-adsorbate system consisting of Co2(CO)8 and HV chamber background hydrocarbons and we show that it allows for a quantitative description of the compositional variations. The CoC nanocomposite material is studied and characterized using a Langevin fit of the Hall signal. We show that this approach yields insight into the nanocrystal mean size, providing a nanocharacterization method of the material. A linear dependency between the electrical resistivity and the Hall resistivity at saturation is found. Empirically, the material is found to exhibit an optimum in terms of magnetic field resolution at around 65 at.% Co, corresponding to a trade-off between magnetic field sensitivity and electrical resistivity. Finally, we demonstrate single superparamagnetic bead detection using a sub-micron, CoC Hall sensor. Using a novel nanomanipulation setup providing a magnetic coil integrated into a SEM, we show the bead tracking capacities of these sensors.