The fast expanding distributed and cloud-based storage systems have escalated the necessity to have strong error-correcting protocols that guarantee data integrity, reliability, and security. The conventional coding schemes are usually limited in scalability and structural efficiency when they are implemented in massive storage settings. The paper presents a new model to build error-correcting codes based on finite geometries, with the help of the combinatorial constructions like projective and affine planes, to become more redundancy and fault-tolerant. The algorithm combines geometry-based code construction with algebraic encoding and decoding, which allows to effectively detect and correct errors in high-noise scenarios. Experimental assessments reveal that the model proposed has an error correction rate of 96.7% which is more reliable than classical Reed-Solomon codes and the error correction rate less than the classical error correction rate. The framework also provides a high level of resiliency to data corruption and provides a scalable storage architecture. These results suggest that finite geometry-based codes have a promising future in building secure, efficient and high-performance data storage systems.