Nowadays, the number of photos is exploding and it is getting harder to find the photo you want from the extremely large photo albums. This trend urgently necessitates efficient technologies to organise and utilise such huge amount of photos for both end users and photo sharing websites. For example, one person who wants to show his cat’s photo to his friends during a party ends up embarrassingly running through lots of irrelevant photos before his friends diminish their interests. Similarly, a photographer who just captures some photos about a landscape and wants to compare with similar photos he previously took, ends up searching inefficiently from massive albums if he did not spend tremendously time to manually organize all the photos. To solve these problems, we present a novel system for smart photo management built upon two core functionality, namely automatic image tagging and relevant photo search. With our system, users can easily find the photo they want by some simple keywords, dates or locations. In addition, it provides an efficient precision search to explore visually similar images, and smart organisation which could automatically index and organise the photos into common searchable categories without manual annotation.