This technical report outlines our submissions to the CHiME-8 NOTSOFAR-1 Challenge single-channel track, which focuses on distant speaker diarization and automatic speech recognition. This track evaluates far-field meeting transcriptions on a single audio channel using a single device. Our submission features a highly streamlined and efficient system incorporating both a non-autoregessive speaker diarization and automatic speech recognition model. Each submitted system outperforms the baseline in terms of Time-Constrained minimum Permutation Word Error Rate (tcpWER) on the development set. We provide an analysis on models sizes and inference throughput under constrained computational resources with the most practical system using less than 100 million parameters. Additionally, we report new state-of-the-art results for the AMI Mix and AMI SDM datasets with DER values of 11.83 % and 17.55 %, respectively.