With the rapid development of the Internet and digital advertising market, the scale of the global advertising market continues to expand, and the user behavior mode has undergone fundamental changes due to the popularity of mobile devices, which puts forward higher requirements for the real-time, personalized and effect tracking of advertising. Traditional advertising delivery systems face core challenges such as real-time data processing delays, inaccurate capture of user interests, and reliance on outdated information for delivery strategies, leading to decision-making errors and resource waste by advertisers. This study is based on a full stack development perspective, following the entire software engineering lifecycle. Through requirement analysis, the system's functional and non-functional boundaries are clarified, and a layered architecture design is adopted to divide advertising promotion management, advertising push, advertising log management, and real-time data warehouse into four modules; Use class diagrams, flowcharts, and sequence diagrams to complete module modelling and code development, and utilize streaming data processing technologies such as Flink and Kafka to achieve real-time streaming data processing and large screen display of advertising effects. Aiming at the optimization problem of advertising ranking model, a DISM model integrating DSSM dual tower model and FM factorization machine is proposed to effectively solve the shortcomings of DIN model in deep level interest mining and low-level feature interaction learning of user behavior sequences, significantly improving CTR and RPM indicators. The system verifies reliability through functional and non-functional testing, meeting the needs of advertisers for ad creation and push, real-time effect analysis, as well as administrator ad approval and log collection. The real-time data warehouse module assists advertisers in adjusting their advertising strategies in real time. This study forms a complete closed loop of "requirements design implementation optimization testing", providing a performance optimization path from the perspective of full stack development for the efficient and intelligent implementation of advertising delivery systems. In the future, it can further deepen the direction of recall algorithms, coarse ranking models, data indicator improvement, and model optimization.

Taige Zhang. Performance Optimization and Implementation Pathways of Advertising Delivery Systems from a Full-Stack Development Perspective. International Journal of Multimedia Computing (2025), Vol. 6, Issue 1: 126-134. https://doi.org/10.38007/IJMC.2025.060112.

[1] Ghadimi M, Baghayi N, Shateri A.DataBay: A Unified Platform for Automating Data Warehouse Management, Real-Time Data Processing, and Ensuring Data Quality and Monitoring[J].  2025.DOI:10.36227/techrxiv.173834980.04708005/v1.

[2] Fan X, Lu J. Enterprise Level Data Warehouse System Based on Hive in Big Data Environment[J].Procedia Computer Science, 2024, 243:67-75.

[3] Zhang T, Jin X, Bai S, et al. Smart Public Transportation Sensing: Enhancing Perception and Data Management for Efficient and Safety Operations[J].Sensors (14248220), 2023, 23(22).DOI:10.3390/s23229228.

[4] Chang, Chen-Wei. "AI-Driven Privacy Audit Automation and Data Provenance Tracking in Large-Scale Systems." (2025).

[5] Li W. Building a Credit Risk Data Management and Analysis System for Financial Markets Based on Blockchain Data Storage and Encryption Technology[C]//2025 3rd International Conference on Data Science and Network Security (ICDSNS). IEEE, 2025: 1-7.

[6] Huang, J. (2025). Balance Model of Resource Management and Customer Service Availability in Cloud Computing Platform. Economics and Management Innovation, 2(4), 39-45.

[7] Zhang K. Research on the Application of Homomorphic Encryption-Based Machine Learning Privacy Protection Technology in Precision Marketing[C]//2025 3rd International Conference on Data Science and Network Security (ICDSNS). IEEE, 2025: 1-6.

[8] Tang X, Wu X, Bao W. Intelligent Prediction-Inventory-Scheduling Closed-Loop Nearshore Supply Chain Decision System[J]. Advances in Management and Intelligent Technologies, 2025, 1(4).

[9] Pinto R C, Tavares A R. PReLU: Yet Another Single-Layer Solution to the XOR Problem[J].  2024.

[10] Xu, H. (2025). Optimization of Packaging Procurement and Supplier Strategy in Global Supply Chain. European Journal of Business, Economics & Management, 1(3), 111-117.

[11] Pan, H. (2025). Development and Optimization of Social Network Systems on Machine Learning. European Journal of AI, Computing & Informatics, 1(2), 73-79.

[12] Wu X, Bao W. Research on the Design of a Blockchain Logistics Information Platform Based on Reputation Proof Consensus Algorithm[J]. Procedia Computer Science, 2025, 262: 973-981.