Penguasaan kosakata bahasa Inggris pada siswa sekolah dasar masih tergolong rendah dan proses pembelajaran yang berlangsung cenderung kurang interaktif sehingga minat siswa dalam pembelajaran kosakata belum optimal. Kondisi tersebut menunjukkan perlunya media pembelajaran yang lebih menarik dan mampu memberikan pengalaman belajar yang lebih variatif. Penelitian ini bertujuan untuk merancang dan membangun game puzzle kosakata bahasa Inggris berbasis gamification dengan menerapkan metode Finite State Machine (FSM) pada AI Agent emosional serta algoritma Fisher–Yates Shuffle untuk menghasilkan variasi soal. Penelitian ini menggunakan metode Game Development Life Cycle (GDLC) yang terdiri dari tahap initiation, pre-production, production, testing alpha, beta release, dan release. Proses pengembangan dilakukan menggunakan Unity 2D dengan pengumpulan data melalui observasi, wawancara, dan studi pustaka. Metode FSM diterapkan untuk mengatur perubahan state karakter sesuai kondisi permainan, sedangkan algoritma Fisher–Yates Shuffle digunakan untuk mengacak urutan soal dan posisi jawaban sehingga permainan menjadi lebih bervariasi. Hasil pengujian Black Box menunjukkan bahwa seluruh fungsi sistem dapat berjalan sesuai rancangan, sedangkan pengujian White Box menunjukkan bahwa implementasi FSM dan Fisher–Yates Shuffle telah berjalan sesuai logika program yang dirancang. Selain itu, hasil User Acceptance Test (UAT) yang dilakukan terhadap 35 siswa kelas VI memperoleh persentase sebesar 94,3% dengan kategori sangat baik. Hasil tersebut menunjukkan bahwa game yang dikembangkan mampu memberikan pengalaman belajar yang lebih interaktif, menarik, dan dapat meningkatkan minat siswa dalam pembelajaran kosakata bahasa Inggris.
English vocabulary mastery among elementary school students remains relatively low, and the learning process tends to be less interactive, resulting in students’ interest in vocabulary learning being less than optimal. These conditions indicate the need for more attractive learning media that can provide a more varied learning experience. This study aims to design and develop a gamification-based English vocabulary puzzle game by implementing the Finite State Machine (FSM) method for the emotional AI Agent and the Fisher–Yates Shuffle algorithm to generate more varied questions. The study employed the Game Development Life Cycle (GDLC) method, which consists of initiation, pre-production, production, testing alpha, beta release, and release stages. The game was developed using Unity 2D, while data collection techniques included observation, interviews, and literature review. The FSM method was applied to regulate character state transitions according to game conditions, whereas the Fisher–Yates Shuffle algorithm was implemented to randomize the sequence of questions and answer positions, thereby creating a more varied gameplay experience. Black Box testing results showed that all system functions operated according to the intended design, while White Box testing demonstrated that the implementation of FSM and Fisher–Yates Shuffle followed the designed program logic. Furthermore, the User Acceptance Test (UAT) conducted on 35 sixth-grade students obtained a score of 94.3%, which falls into the very good category. These results indicate that the developed game provides a more interactive and engaging learning experience and can increase students’ interest in learning English vocabulary.