<![CDATA[Abstrak Transformasi digital pendidikan menuntut guru memiliki literasi teknologi dan kemampuan pedagogik berbasis data, namun sebagian guru ISMUBA masih memiliki pemahaman terbatas mengenai pemanfaatan Deep Learning dalam pembelajaran. Pengabdian ini bertujuan meningkatkan kompetensi pengetahuan dan keterampilan guru dalam merancang serta mengimplementasikan pembelajaran berbasis Deep Learning. Metode pelaksanaan menggunakan pendekatan partisipatif melalui sosialisasi, pelatihan konseptual dan praktik, pendampingan project-based learning, serta evaluasi. Mitra kegiatan adalah 15 sekolah Muhammadiyah di Surakarta dengan 60 guru ISMUBA sebagai peserta. Evaluasi dilakukan menggunakan pre-test dan post-test serta penilaian produk pembelajaran. Hasil menunjukkan peningkatan kompetensi yang signifikan, yaitu kategori pengetahuan tinggi meningkat dari 25% menjadi 70%. Selain itu, peserta berhasil mengembangkan minimal tiga model pembelajaran berbasis Deep Learning yang siap diterapkan serta meningkatkan kepercayaan diri guru dalam penggunaan teknologi pembelajaran. Program juga membentuk komunitas kolaboratif guru berbasis teknologi sehingga berpotensi meningkatkan kualitas pembelajaran adaptif dan berbasis data secara berkelanjutan. Kata kunci: deep learning; literasi digital; ISMUBA; pembelajaran berbasis data; pengabdian masyarakat. Abstract The digital transformation of education requires teachers to have technological literacy and data-based pedagogic skills, but some ISMUBA teachers still have a limited understanding of the use of Deep Learning in learning. This service aims to improve teachers' knowledge competencies and skills in designing and implementing Deep Learning-based learning. The implementation method uses a participatory approach through socialization, conceptual and practical training, project-based learning assistance, and evaluation. The partners of the activity were 15 Muhammadiyah schools in Surakarta with 60 ISMUBA teachers as participants. Evaluation is carried out using pre-test and post-test as well as assessment of learning products. The results showed a significant increase in competence, i.e. the category of high knowledge increased from 25% to 70%. In addition, participants managed to develop at least three Deep Learning-based learning models that are ready to be implemented and increase teachers' confidence in the use of learning technology. The program also forms a collaborative community of technology-based teachers so that it has the potential to improve the quality of adaptive and data-based learning in an ongoing manner. Keywords: deep learning; digital literacy; ISMUBA; data-driven learning; community service.]]>
