The Problem
Office of the Coordinator-General (OCG) in the Queensland Department of State Development and Infrastructure assesses the impact of significant and complex projects. The assessment process includes consultation which encompasses processing submissions from government agencies, organisations and members of the public, in a variety of document formats. Each submission is required to be considered and consolidated as a public record.
Consolidation, identification and classification of issues found in each submission is a laborious and time-consuming exercise using significant OCG Resources.
The OCG team acknowledged that the maturing of AI capabilities in the Large Language model space presents great opportunities for enhancing efficiency in that process.
The Solution
During the proof of concept (POC) phase, the exposé team worked closely with the OCG to build a pipeline to ingest submissions in diverse formats, extract essential information, and efficiently categorise them using the explAIn engine.
Leveraging exposé’s explAIn engine facilitated a notable acceleration in the development of the solution and enabled the exposé team to deliver consolidated submissions in the format recommended by the OCG team.
Business Benefit
Based on analysis of approx. 600 submissions for a single consultation:
- Reduced processing time from 400 hours to just 20 minutes.
- Labour cost reduction from $30k to just $9(Azure costs).
- Maintained accuracy and consistency in processing submissions while identifying further optimisation opportunities.
- Reliable conversion of unstructured submissions data into categories for future integration into reporting systems.
- Provided the OCG team with opportunities to redirect focus from tedious tasks to innovative improvements.
Download our full case study here.