Zero Cost Approach for NLP Based, Serverless Voicemail Monitoring Automation Pipeline

This paper introduces a novel and cost-effective approach to voicemail automation by designing a zero-cost approach processing pipeline that leverages Google Voice, Google Cloud's free-tier services, and open-source natural language processing (NLP) models. The proposed system automates the end-to-end workflow starting from transcription parsing to structured response generation making it ideally suited for individual users and small businesses with limited technical or financial resources. Unlike previous solutions that rely on paid NLP APIs or proprietary infrastructure, this system integrates a custom FastAPI microservice that uses the Sentence Transformer model for intent classification and spaCy's for real-time entity extraction. These components are deployed using Google Cloud Run, ensuring scalability and low-latency inference while remaining entirely within free-tier usage limits. The modular, serverless design incorporates Google Apps Script, Cloud Functions, Cloud Storage, and Gmail SMTP, enabling eventdriven automation for handling voicemail actions such as appointment scheduling, and general inquiries. By eliminating cloud infrastructure costs and minimizing setup complexity, this work demonstrates the practical feasibility of enterprise-grade voicemail automation accessible to any user. The architecture aligns with recent advancements in lightweight NLP, triggerbased workflows, and zero-cost cloud deployment strategies, contributing to the growing body of research on real-time, serverless business process automation.