Service Needs

Leveraging Artificial Intelligence (AI) for Automated Processing of Licence Applications

- Manual processing of licence applications involves intensive checking of application forms and supporting documents, leading to long processing time and heavy workload on licensing staff. - Unstructured documents in formats such as PDF, JPG and PNG make it difficult to extract key data consistently for vetting and data entry into existing licensing systems. - The current process is prone to human errors in data verification and cross-checking against licensing criteria and records. - Some applications are submitted in paper application forms, and attached with various supporting documents, staff must scan and file documents one by one and then manually enter data into the system, making the workflow cumbersome. - Due to the variety of data types and inconsistent formats, the department has to follow up with applicants in writing or by phone to amend details and request supplementary documents; the assessment and referral process also involves multiple manual checks, returns/supplement requests, and subsequent reviews/referrals by staff of different ranks, which can easily lead to case backlog during peak application periods