About Me

My name is Hao Hao Tan (郑豪好), or you can call me Harry. I am from Kuala Lumpur, Malaysia, and currently residing in Singapore.

I currently work on AI-powered features at BandLab to enhance music creation workflow, e.g. SongStarter, Voice Changer, Splitter, MIDI tools, Audio-to-MIDI, Similar Sounds, etc. I also work on on-device model serving and optimization – some of the features above run directly on users’ mobile device!

I’m passionate about building technologies that enhance creative applications. My bachelor’s thesis was about a chord-based music generation system using LSTMs. Then, I worked on controllable music generation and audio synthesis at Prof. Dorien Herremans’ lab. I later joined Fairphonic and built a deep learning–based music content ID system (originals and covers included!) to fix issues on rights attribution. Before joining BandLab, I did backend engineering at TikTok on its (massive) feed recommendation system.

Sometimes I work on independent research / open-source projects, you can find them on my GitHub page. I’m generally interested about topics related to audio, synthesizers, (fun ways of) search & retrieval, and small on-device models.

I am a classically trained pianist, and I listen to all kinds of music ranging from Cantonese classics, Edvard Grieg to pure electronic beats. Stay tuned for some awesome playlists that I made.

Feel free reach me at helloharry66 [at] gmail or @GoodGood014 on Twitter.