Automate Onboarding Forms with Intelligent Document Processing
The Leobit team developed a PoC for a solution that extracts information from PDF documents and automatically populates the required onboarding forms. The tool also supports voice commands, allowing users to specify information and add new fields to the forms. It also tracks the origin of every field value for greater transparency.
The solution’s front end is built with Angular and Tailwind CSS, while .NET is the primary back-end technology. Azure Document Intelligence 4.1 handles text extraction from uploaded documents, while Azure OpenAI processes the extracted text through a structured prompt to identify and infer onboarding fields. The ElevenLabs AI model allows users to refine form data with voice commands.
The framework can be extended to any document-to-form workflow where data needs to be extracted from unstructured inputs.