With the introduction of the concept of wisdom education, wisdom education has become the trend and direction of the future development of education in China, and wisdom education has become one of the important means to promote educational reform and improve the quality of education. In the current field of higher education, due to the complexity of information and the endless emergence of various professional competitions, students are faced with the problem that they cannot efficiently and accurately choose competitions suitable for their professional development to meet their own personalized education and in terms of information acquisition and utilization. Different students have great information difference and cannot get professional suggestions that are really effective and suitable for their own development. Not only that, most schools are also faced with the dilemma that counselors are non-professional practitioners, resulting in the counselors who communicate with students the most are unable to give relevant professional advice, while professional teachers with relevant professional knowledge and abilities are not familiar with students and cannot form effective guidance to students. In order to promote the construction and development of smart education, and comprehensively use modern information technology to integrate various resources in the process of education and teaching, realize the effective integration of teaching, management and service, and promote the all-round development of students and the improvement of teachers' teaching ability, the team decided to develop this work, using students' resume information, competition experience, skill level and other aspects, combined with the current industry development trend, advise students on career paths appropriate to their individual circumstances. At the same time, counselors and professional teachers are added to solve the pain points that counselors do not know the direction of students' learning and cannot give more reasonable suggestions, and professional teachers and students do not understand each other and cannot provide professional guidance due to lack of contact. This work uses deep learning OCR field technology with NER field technology to build the algorithm model. First, the text information in the student's resume is extracted using the DB algorithm in the OCR field with the CRNN algorithm, and then the extracted information is marked and preprocessed using sklearn and pandas. The preprocessed data is fed into the GlobalPointer neural network model to extract the relevant entities. Combined with the Word2vec and Doc2vec model to obtain the relevant embedding of entity and competition description, the matching degree evaluation of students participating in the competition is completed by calculating the cosine similarity, and finally personalized suggestions and guidance are provided. In terms of visualization and front-end development, the React framework Next.js and Antd component libraries are used for development, and state management is carried out through React-redux. Data storage and back-end development are developed using the SpringBoot MyBatisPlus framework, combined with MySQL database for data persistence operations. The development of this work can solve all the pain points and difficulties mentioned above, really help the development and construction of smart education, provide students with personalized career development suggestions, help students to formulate their own career planning direction more accurately, and enhance their career competitiveness and development potential This work includes the following main functions according to the different user subjects: ① Students: Get the official website links of each competition through this work, and at the same time understand the fit between each competition and oneself, and get the ability improvement and suggestions for oneself through this work. ② Counselor: Through this work, we can understand the overall fit of a certain competition student, and can screen out the student list according to the custom grade, at the same time, we can obtain the resume rating of the guided students, and give the main improvement points. ③ Professional teachers: Through this work, you can check the voluntary guidance information for a related competition, and you can match the relevant competition guidance award information according to the award certificate uploaded by the students. In order to realize the above functions, the algorithm direction of this work decided to use OCR in the field of deep learning with NER technology. First, the text information in the student's resume is extracted using the DB algorithm in OCR technology with the CRNN algorithm, and then the extracted information is preprocessed using sklearn and pandas. The preprocessed data is fed into the GlobalPointer neural network model to extract the relevant entities and obtain the relevant entities. The related entity is operated on after the related entity is obtained. The front end of the development direction uses the React framework Next.js and the Antd component library for development, and the state management is carried out through the React-redux. The back-end uses the Java SpringBoot MyBatisPlus framework for development, combined with the MySQL database for data persistence operations.