In this paper, we design an English writing teaching platform based on the English automatic error correction model as a teaching strategy for English writing for college students, with a view to realizing the improvement of students’ English writing efficiency. The English automatic error correction model is based on the encoder-decoder framework and combines the hybrid attention mechanism to improve the error correction performance of the model. The syntactic encoder and semantic encoder use Bi-GRU and BERT to extract information, and the decoder side uses the hybrid attention mechanism to improve the accuracy of decoding. The English writing teaching platform is constructed from functional modules such as text error correction module, data collection module and background management module. The English writing teaching platform in this paper has the fastest reduction in the loss function, which always fluctuates in the range of 40~60 after 80K steps of training, while the F0.5 value is 56.34, and the P value and R value are 66.84 and 35.11, respectively, which is superior to other comparative models in terms of the loss function results and accuracy performance. The experimental classes applying this paper’s platform for teaching English writing showed significant differences (P<0.05) in the mean values of the five writing competence dimensions of content expression, organization, vocabulary use, grammar use, and standardized writing, which were higher by 2.53, 0.91, 0.91, 1.18, and 0.52, respectively.