From February to July 2022, Archives New Zealand led a proof-of-concept project to explore machine learning (ML) and hyperscale cloud computing for auto-classifying digital public records, with a focus on surfacing information relevant to Māori. Funded by the Digital Government Partnership Innovation Fund, the project aimed to tackle challenges in managing vast digital information from public offices, much of which holds long-term archival significance under the Public Records Act 2005. Traditional systems cannot manage the massive volume of digital records, making manual sorting impractical. Archives New Zealand collaborated with stakeholders, including the Ministry of Justice, Ministry for Primary Industries, Microsoft, and AWS, to evaluate ML tools for appraising records and identifying Māori-specific content. Initial results were promising, but further model refinement is needed. Key future steps include ongoing collaboration with Māori, ensuring cultural relevance, and aligning with the Algorithm Charter. This project highlights the transformative potential of ML in digital record management, setting the stage for more efficient government archival practices and culturally inclusive records management.