Worked on the apache/tika repository to address a MIME type detection issue affecting HTML content with an iframe as the root element. Focused on improving the accuracy of content classification, the developer implemented a targeted fix in Java to ensure that HTML snippets beginning with an iframe are correctly identified as text/html rather than misclassified as application/xml. The solution included adding a regression test to maintain reliability and prevent future recurrence of the issue. This work demonstrated skills in MIME type detection and testing, contributing to more robust handling of malformed or partial HTML content in streaming and embeddable scenarios.