“Hanoi” Hantrakul
ลำธาร หาญตระกูล
Cultural Technologist,
AI Research Scientist,
Composer
TIKTOK / BYTEDANCE (字节跳动) SEPTEMBER 2020—Present
SENIOR RESEARCH SCIENTIST
Speech Audio and Music Intelligence Team (SAMI)
- Developed AI-powered audio processing for multiple platforms including TikTok, CapCut and SoundOn.
- Co-inventor of the realtime audio ML algorithm in Mawf and TikTok on-device vocal effects. 5 company patents.
- Point of contact for cross-functional projects combining music, technology, business and product strategy e.g. Ripple App
Mountain View, USA
Bangkok, Thailand
Shanghai, China
GOOGLEJUNE 2018—SEPTEMBER 2020
AI RESIDENT
- Co-author of Google Magenta’s open-source Differentiable Digital Signal Processing (DDSP) library
- Tech lead behind audio synthesis stack used in Tone Transfer and Sounds of India Google products.
- Published papers on neural audio synthesis in ICLR, ICML, ISMIR
San Francisco, CA, USA
SMARTEARAUGUST 2017—JUNE 2018
AUDIO DEEP LEARNING ENGINEER
- Implemented company’s core technology for ML-based Auditory Scene Analysis (ASA)
San Francisco, CA, USA
CREATIVE EXPERIENCE (MUSIC):
YABOI HANOIMAY 2021—Present
Composer /Sound Design
- Pichet Klunchun (Taiwan International Arts Festival 2024)
- Saran Yen Panya (Cartier Collection 2023)
- Pinaree Sanpitak (Bangkok Arts Biennale 2022)
- Goethe Institute (Common Tonalities 2021)
Creative Director
- Commissions by the Thai Ministry of Commerce (Thailand Gems and Jewels International Convention 2022)
Remote
PUBLICATIONS (SELECTED):
Differentiable Wavetable Synthesis.
Shan S., Hantrakul L., Chen J., Trevleyan D., Avent M. ICASSP 2022
Differentiable Digital Signal Processing.
Engel J., Hantrakul L., Gu C., Roberts A. ICLR 2020
Fast and Flexible Neural Audio Synthesis
Hantrakul L., Engel J., Roberts A., Gu C. ISMIR 2019
Interactive Neural Audio Synthesis
Hantrakul L., Engel J., Gu C., Roberts A. ICML Workshop 2019
PATENTS (SELECTED):
CN Patent 202211289254.0 “BI-DIRECTIONAL MUSIC PROCESSING SYSTEM FOR INTELLIGENT ASSISTED COMPOSITION”
US Patent 17/748,882“METHOD AND SYSTEM FOR REAL-TIME AND LOW LATENCY SYNTHESIS OF AUDIO USING NEURAL NETWORKS AND DIFFERENTIABLE DIGITAL SIGNAL PROCESSORS”
US Patent 11,735,197“MACHINE LEARNED DIFFERENTIABLE DIGITAL SIGNAL PROCESSING”
US Patent 17/525,814“DIFFERENTIABLE WAVETABLE SYNTHESIZER”
INTERNATIONAL AWARDS (SELECTED):
AI Song Contest 2022
1st Place Winner for Enter Demons and Gods
A'Design Competition 2017
Musical Instruments category: Silver Award (Top 5%) for fidular
Core77 Design Competition 2017Open Design category: Student Runner Up for fidular
EDUCATION:
GEORGIA INSTITUTE OF TECHNOLOGYAUGUST 2016—MAY 2018,
M.S Music Technology (ML and DSP for audio processing)
- Graduate Research Assistant in Robotic Musicianship Group
- Thesis: Regressing finger flexions using machine learning and single-element transducers. Advisors: G. Weinberg, A. Lerch, B. Boots
Atlanta, GA, USA
YALE UNIVERSITYAUGUST 2011—MAY 2015, Cum Laude (Top 15%),
B.S Applied Physics (Distinction in the Major)
- Thesis: Surrogate Soundboards for novel transmission of instrumental audio. Advisors: R. Kuc, L. Willen
B.A Music (Distinction in the Major)
- Thesis: The arrangement of Thai Traditional music for non-Thai ensembles. Advisors: D. Harrison
New Haven, CT, USA
LANGUAGES
English (Native)
Thai (Native)
French (Business proficiency. TCF B2 Level.)
Mandarin (Preparing for 6/2024 HSK5)
Python (Pytorch, Tensorflow | ML frameworks)
C++ (JUCE | Realtime audio frameworks)
Comprehensive Academic CV available upon request.