AI & Creativity

As First MIR Engineer

Research, prototyping and implementation of Search engines, Recommendation Systems and Machine Learning algorithms for Music Discovery. Also worked on signal processing and audio fingerprinting as well as Women in Tech outreach programmes.

• Who: Amélie Anglade
• Employer: SoundCloud
• Date: July 2011 - August 2013
• Domain: Music Information Retrieval, IR and Machine Learning



• Watch: DiscoRank presentation at the Realtime Conf 2013

Close Window

As Senior Engineer

In charge of building recommendation/personalisation features as well as the entire web backend development. Leading the developer team, onboarding, training and supervising programmers across the entire stack.

• Who: Amélie Anglade
• Employer: frestyl
• Date: September 2013 - June 2014
• Domain: RecSys and Web Development



• Watch: frestyl panel at the Berlin Apple Store

Close Window

As Graduate Student Researcher

Research and implementation of clustering techniques to create and maintain user communities that share specific music tastes in order to build P2P radio channels.

• Who: Amélie Anglade
• Affiliation: Philips Research
• Date: August 2006 - January 2007
• Domain: RecSys, Complex Network Theory, Clustering Algorithms, P2P



• Read: Publication at RecSys 2007

Close Window

As Principal Data Scientist

Performed product-focused Data Science tasks for the mobile application Bertha, and surrounding services. Took care of the entire Data Science stack: designing, prototyping, testing and scaling Data Science algorithms, including Big Data Engineerings tasks. Collaborated with backend engineers, dev ops, designers and product owners to fully integrate Data Science into their code, work and processes while applying an agile methodology to the Data Science processes as well, including quick iterations and bi-weekly demos of data science results. Trained colleagues, supervised working students, recruited and onboarded data scientists.

• Who: Amélie Anglade
• Where: Mobile services team within Mercedes-Benz.io
• Date: February 2017 - September 2018
• Domain: Data Science, Time Series Analysis, Predictive Modelling



• Check out: The Bertha App

Close Window

As Postgraduate Researcher

PhD candidate, affiliated with the Centre for Digital Music and the OMRAS2 project (large UK project funded by the Engineering and Physic Sciences Research Council). Conducted research and development on knowledge representation of musical harmony and modelling of genre, composer and musical style using Machine Learning techniques and Logic Programming.

• Who: Amélie Anglade
• Affiliation: Centre for Digital Music, Queen Mary University of London
• Date: April 2007 - June 2011
• Domain: Research, MIR, harmony, Automatic Genre Classification, Inductive Logic Programming



• Read: PhD thesis

Close Window

As Co-founder and Chief Product Officer

Think of Melodrive as as a virtual composer sitting inside a digital world. The composer listens to the events happening in an experience and creates music that matches them. On the fly.

For example, when a player gets a hit from an enemy the music can respond by being emotionally more angry or dramatic. This real-time generation process is possible thanks to advancements in AI techniques, patented by Melodrive, which enable flexible change to the music to suit any emotional setting.

• Who: Ryan Groves
• Affiliation: Melodrive
• Date: Oct 2016 - Feb 2019
• Domain: R&D, Fundraising, Product Management, Planning, Automatic Composition



• Watch: AI Composition Plug-in

Close Window

As Music Algorithms Developer

Zya enables anyone with a PC, literally, to easily create high-quality music. People can start creating a song in any genre by using their own voice to hum a beat or baseline. They can also tap or touch to create the beat. After inputing their own music, users can mix in hooks, some of which are from hit songs. At launch Zya has Lady Gaga's "Poker Face" and "Just Dance," as well as pop hits like "Like a Virgin," "Born to Be Wild" and "Bye Bye Bye." Zya's technology then adjusts the various parts of the song to make them fit together.

• Who: Ryan Groves
• Affiliation: Music Mastermind (now Zya)
• Date: May 2010 - June 2012
• Domain: R&D, Vocal Transcription, Automatic Key Detection, Automatic Accompaniment



• Watch: Assisted Composition App Demo

Close Window

As Director Of Research and Development

Ditty is an automated musical messenger. You text. Ditty sings! Make every message a hit.

#1 Music App in 25 countries

2015 Music App of the Year! – Silicon Beach Award

2015 Music App of the Year! -Appy Awards

Top Ten “Apps of the Year”- Apple ‘n’ Apps

• Who: Ryan Groves
• Affiliation: Zya
• Date: Aug 2014 - July 2015
• Domain: R&D, Text-to-Melody Alignment, Melodic Embellishment, Melodic Phrase Detection, Audio Streaming



• Watch: Ditty Trailer

Close Window