• Kheni Manav

Organizations have started moving their Large Language Model services from cloud based operations to on premises systems, which now require fast and secure processing of sensitive data. The research paper studies how local and cloud based LLM systems perform through their effects on system response times, data processing rates, and their ability to protect user information. The system Edge AI pipeline operates through a PyQt5 desktop application suite, which uses Ollama to perform inference processing on local systems. The system received deployment at Cecil Labs to support their regulatory operations, which need complete adherence to GDPR and HIPAA standards. Our results demonstrate that localized LLMs effectively eliminate network induced latency while maintaining 100% internal data residency. The system implementation brought down manual research work by 60% and simultaneously boosted total operational performance by 35%. The research shows that organizations can run their operations on normal business computers through 4 bit quantization and OCR based digitization. This provides a secure environment for processing sensitive institutional data instead of using cloud based APIs.

Edge AI, Ollama, data sovereignty, LLM inference, privacy preserving AI

"Bridging the Privacy Gap: A Localized Edge AI Pipeline for Regulatory Compliance", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 2, page no.c615-c618, February-2026, Available :http://www.jetir.org/papers/JETIR2602292.pdf

"Bridging the Privacy Gap: A Localized Edge AI Pipeline for Regulatory Compliance", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 2, page no. ppc615-c618, February-2026, Available at : http://www.jetir.org/papers/JETIR2602292.pdf