A Scalable AI Framework for SMEs and a Novel Metric: Proactive Adaptation Rate (PAR)

Abstract

This paper presents Kolay.AI, a scalable AI framework designed to address the key challenges faced by SMEs, including cold starts, model drift, data privacy, and limited computational resources. Its novel contribution is the Proactive Adaptation Rate (PAR) metric, which enables AI systems to anticipate market shifts and make preemptive adjustments. The framework leverages FFT, LSTM, ARIMA, and Linear Regression models on real-world data from 40 SMEs across diverse sectors. Through federated learning and modular service architecture, it delivers secure, high-performance AI solutions without compromising data privacy. Evaluation metrics (RMSE, MAE, MAPE, PAR) demonstrate significant improvements in forecasting accuracy and operational efficiency. Results show that SMEs using Kolay.AI achieved 15% 20% better cash flow, reduced operational downtime, and maintained alignment with dynamic market conditions. The PAR metric proved critical for early anomaly detection, enabling optimised inventory management and risk mitigation. By shifting the focus from reactive to proactive AI evaluation, this work advances SME-centric AI adoption. The framework’s integration of federated learning and PAR-driven monitoring sets a foundation for future research on adaptive AI in resource-constrained environments.

Keywords

• AI Forecasting
• SMEs
• FFT
• LSTM
• CRM

References

1. Seker, Sadi Evren. "KOBİ’lere Özel Basit Yapay Zeka Çözümü: Kolay. AI." Yönetim Bilişim Sistemleri Ansiklopedi (2023): v.11, is. 1, pp: 24 google scholar
2. Yoruk, Rabia. "Enhancing SME Operations with Machine Learning and Business Intelligence: A Case Study of Kolay. ai." Journal of Data Analytics and Artificial Intelligence Applications 1.1 (2025): 61-83. google scholar
3. Fatoki, O. (2012). An investigation into the financial management practices of small and medium enterprises in South Africa. Jour- nal of Social Sciences, 33(2), 179-188. google scholar
4. Turgut, G., Cheruiyot, P. K., & Sang, H. W. (2021). Effect of cash flow forecasting on financial sustainability of SMEs in Kericho Central Business District. International Journal of Scientific and Research, 11(10), 103–107. google scholar
5. Haataja, T. (2016). Sales forecasting in small and medium-sized enterprises. Helsinki Metropolia University of Applied Sciences google scholar
6. Chong, F., Aulbach, S., & Jacobs, D. (2006). A survey of data management in SaaS architectures. Datenbanksysteme in Business, Technologie und Web (BTW), 408-422. google scholar
7. Menzel, M., & Ranjan, R. (2012). A survey of security in multi-tenant cloud computing. In 2012 IEEE 8th World Congress on Services (pp. 377-384). google scholar
8. Kairouz, P., McMahan, H. B., Avent, B., Bellet, A., Bennis, M., Bhagoji, A. N.. & Zhao, S. (2021). Advances and open problems in federated learning. Foundations and Trends in Machine Learning, 14(1–2), 1-210. google scholar

9. Li, T., Sahu, A. K., Talwalkar, A., & Smith, V. (2020a). Federated learning: Challenges, methods, and future directions. IEEE Signal Processing Magazine, 37(3), 50-60 google scholar
10. McMahan, H. B., Moore, E., Ramage, D., Hampson, S., & y Arcas, B. A. (2017). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics (pp. 1273-1282). PMLR. google scholar
11. Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M., & Bouchachia, A. (2014). A survey on concept drift adaptation. ACM Computing Surveys (CSUR), 46(4), 1-37. google scholar
12. Schein, A. I., Popescul, A., Ungar, L. H., & Pennock, D. M. (2002). Methods and metrics for cold-start recommendations. In Proceedings of the 25th annual international ACM SIGIR conference on Research and development in information retrieval (pp. 253-260). google scholar
13. Lu, J., Liu, A., Dong, F., Gu, F., Gama, J., & Zhang, G. (2018). Learning under concept drift: A review. IEEE Transactions on Knowledge and Data Engineering, 31(12), 2346-2363. google scholar
14. Zhang, Z, Zhao, K., Jia, Q., Fang, Y., & Yu, Q. (2020). Large-scale uncertainty estimation and its application in revenue forecasting for SMEs. arXiv – CS - Machine Learning. google scholar
15. Yoshino, N., & Taghizadeh-Hesary, F. (2016). Major challenges facing small and medium-sized enterprises in Asia and solutions for mitigating them. Asian Development Bank Institute Working Paper 564. google scholar
16. Rodriguez, A., Goodwin, E., & Melancon, M. V. (2013). Planning and forecasting: Findings from small and medium-sized enterprises. International Journal of Management and Human Resources, 1(1), 61. google scholar
17. Brijal, P., Enow, S., & Isaacs, B. H. (2014). The use of financial management practices by small, medium, and micro enterprises: A perspective from South Africa. Industry & Higher Education, 28(5), 341–350. google scholar
18. Filipe, S. F., Grammatikos, T., & Michala, D. (2016). Forecasting distress in European SME portfolios. Journal of Banking & Finance, 64, 112–135. google scholar
19. Yong et al. (2019). A review of recurrent neural networks: LSTM cells and network architectures. Neural Computation, 31(7), 1235– 1270. google scholar
20. Schalck, C., & Yankol-Schalck, M. (2021). Predicting French SME failures: New evidence from machine learning techniques. Applied Economics, 53(51), 5948–5963. google scholar
21. Musbah, H., & El-Hawary, M. (2019). SARIMA model forecasting of short-term electrical load data augmented by fast Fourier transform seasonality detection. IEEE Canadian Conference of Electrical and Computer Engineering (CCECE). google scholar
22. Zhu, Y., Li, X., Wang, W., & Li, H. (2019). Factors influencing the adoption of artificial intelligence by small and medium-sized enterprises in China. Sustainability, 11(18), 4937. google scholar
23. Met, I., Erkoç, A., & Seker, S. E. (2022). Performance, Efficiency, and Target Setting for Bank Branches: Time Series with Automated Machine Learning. IEEE Access, 10, 133463-133474. google scholar