Nuclear reactions provide a powerful tool for investigating the structure and dynamics of atomic nuclei. High-resolution γ-ray spectroscopy of the reaction products provides essential information on their excited-state structure and electromagnetic properties. A particular experimental challenge arises when the nuclei produced in these reactions are moving with substantial velocities. The γ-rays emitted from the moving nuclei undergo a Doppler shift, and the measured energy therefore depends on the angle at which the γ-ray is emitted relative to the direction of motion of the nucleus. Accurate determination of this angle is consequently essential for Doppler correction and for resolving closely spaced nuclear levels. This has motivated the development of highly segmented and γ-ray tracking detector arrays, such as GRETINA and AGATA, where the interaction points of γ-rays are reconstructed to determine their trajectories. In particular, the position of the first interaction point has a strong influence on the Doppler-correction capability of the detector. The development of such tracking detectors places stringent requirements on the determination of γ-ray interaction positions. At the same time, increasing detector granularity leads to a large number of readout channels and greater electronic and data-acquisition complexity. This raises the question of whether precise γ-ray interaction positions can be obtained using a compact detector architecture while keeping the readout complexity manageable. Position reconstruction from the SiPM response has been investigated using conventional signal-processing techniques as well as machine-learning-based approaches. The developed detector demonstrates approximately millimetre-scale spatial localization and provides a compact approach to position-sensitive γ-ray detection. This work explores the potential of combining monolithic scintillators, sparse SiPM readout, and advanced position-reconstruction methods to obtain precise spatial information with reduced readout complexity. The developed detector can serve as a tool for localized response mapping and calibration of segmented γ-ray detector systems, and may find applications in future nuclear-reaction experiments, such as those planned with tracking arrays like AGATA and GRETINA, where accurate γ-ray interaction-position information is essential for improving the reconstruction of reaction observables.