Architecting Multi-Cloud Salesforce Integrations for Scalable Enterprise Applications

Enterprise operations are becoming increasingly complicated, calling for smart, real-time orchestration across several cloud platforms such as AWS, Azure, and GCP. This article presents an innovative AI-driven architecture called "architecting multi-cloud salesforce integrations for scalable en...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2025 International Conference on Innovations in Intelligent Systems: Advancements in Computing, Communication, and Cybersecurity (ISAC3) s. 1 - 6
Hlavní autoři: Adi, Mahesh, Jha, Debashis, Aturu, Reddy Prasanna, Bethi, Sushmithareddy, Bommana, Ranjith, Khayrullayev, Ismatulla
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 25.07.2025
Témata:
Computational redundancy
Computer architecture
Convergence
Convolutional neural networks
Enterprise operations
Gated recurrent unit
Gated recurrent units
Long short term memory
Multi-Cloud Salesforce Integrations
Optimization
Painting Training Based optimization
Redundancy
Stability analysis
Technological innovation
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Enterprise operations are becoming increasingly complicated, calling for smart, real-time orchestration across several cloud platforms such as AWS, Azure, and GCP. This article presents an innovative AI-driven architecture called "architecting multi-cloud salesforce integrations for scalable enterprise applications using hybrid long short-term memory (LSTM), gated recurrent unit (GRU), one-dimensional convolutional neural network (1D-CNN) and it is named as LSTM-GRU-1DCNN and Painting training-based optimization. To ensure the stability of temporal patterns, the suggested pipeline starts by ingesting time-series data from distributed salesforce APIs and then applies extensive preprocessing. The API latency, error rate, and traffic pattern forecasting is done using a hybrid deep learning model that uses LSTM, GRU, and 1D-CNN layers. While reducing computational redundancy and extracting localized signal properties, the hybrid stack captures long-range dependencies. Drawing inspiration from the iterative process of human artistic refinement, a painting training-based optimization algorithm is introduced as a fundamental innovation. Compared to more conventional approaches, this one increase model convergence stability by adaptively tuning hyperparameters. The experimental results show that the suggested model outperforms the baselines of SVM, KNN, ELM, GRU, 1DCNN, and CNN+LSTM, achieving an accuracy of 97.8%. Optimization benchmarks show that, in comparison to PSO, GWO, and ACO, among others, there is less variance and faster convergence. By lowering API retry rates to 2.4%, lowering response latency by 38%, and increasing resource utilization by 20%, the methodology greatly enhances Salesforce orchestration. Additionally, it accomplishes good generalization in 10× traffic situations, guaranteeing scalable deployment. When it comes to AI-enabled enterprise integration and predictive cloud orchestration, this architecture is a giant leap forward.
DOI:10.1109/ISAC364032.2025.11156665