The production of CAR-modified NK cells under GMP conditions for allogeneic immunotherapy remains challenging due to the need for generating multiple doses. Factors such as NK subpopulation, stimulation method, culture system, and transduction strategy influence manufacturing. We aimed to establish a scalable, GMP-compliant platform optimized in a closed system with high yield and efficiency. NK cells were isolated via CD3⁺ depletion (CD3⁻) and optionally enriched for CD56⁺ (CD3⁻CD56⁺), stimulated weekly with CSTX002 feeder cells and cultured in AIM-V culture medium with IL-2. Different seeding densities (S1, S2, and S3) were tested using the G-Rex system. On day 20, NKG2A and NKG2D expressions were analyzed. A CAR.CD19.IL-15/IL-15Rα construct was used for transduction with BaEV or VSV-G envelopes, incorporating a cationic peptide for enhanced efficiency. Transduction efficiency was evaluated at different time points during expansion (T1, T2, and T3), with and without centrifugation. S2 achieved the highest expansion, with fold-changes of 1441 (CD3⁻) and 693 (CD3⁻CD56⁺), while S1 and S3 showed lower values. CD3⁻ cells exhibited higher proliferation than CD3⁻CD56⁺ cells. Using the VSV-G envelope without centrifugation, CAR expression in CD3⁻ cells increased progressively (5%, 14%, 42% at T1, T2, and T3). CD3⁻CD56⁺ cells had higher transduction at T1 (20.5%) and T2 (31%) but declined at T3 (5%). Centrifugation did not enhance transduction efficiency. The BaEV envelope was less efficient than VSV-G, requiring centrifugation, with a maximum efficiency of 40%. Overall, transduction was most effective at T2 and T3, with CD3⁻CD56⁺ cells transduced with VSV-G showing the highest efficiency. Cationic peptides bypassed centrifugation, addressing GMP challenges. Further studies are needed to optimize reproducibility and improve NK-CAR production for clinical applications. Financial support: FAPESP 2013/08135-2, 2020/07055-9, 2024/01305-4, CNPq 442484/2020-8 and FINEP.