AI-Based Voice to Form Filling System: A Review of Speech Recognition, Information Extraction and Intelligent Form Automation

Digital forms support registrations, applications, service requests, surveys, and institutional processes, but completing them can remain difficult for people who face typing barriers, limited digital literacy, language differences, or accessibility challenges. An AI-based voice-to-form filling system aims to make data entry more natural by allowing a user to speak information and converting it into correctly structured form values. This review examines the technologies relevant to the project: automatic speech recognition (ASR), speech-to-text conversion, natural-language processing (NLP), information extraction, semantic field mapping, validation, multilingual interaction, and user confirmation. It considers how voice-driven form automation can coordinate these components and how optical character recognition (OCR) may support document-assisted entry. The discussion emphasizes that transcription accuracy alone does not guarantee successful form completion. Correct extraction, normalization, mapping, correction, privacy, and usability must also be addressed. The review identifies research and implementation priorities for the project, including handling accents and noisy environments, reducing errors that propagate between stages, supporting multiple languages, and providing a secure human-in-the-loop workflow.