Experience

1. 

   Senior Machine Learning Developer 2020 – present

   Unity Technologies · Copenhagen, Denmark

   AI tools for game developers across multiple product areas. Built gameplay simulation agents for modelling player behaviour in game economies, and researched diffusion models for visual asset generation. Worked on ads pipeline monitoring before joining the Unity Muse team — an LLM-driven AI assistant integrated into the Unity Editor — in March 2023.

2. 

   Senior Researcher 2018–2020

   Moodagent · Copenhagen, Denmark

   Music technology research focused on personalisation. Developed music recommendation systems using collaborative filtering based on user interaction patterns with the Moodagent streaming app.

3. 

   Researcher 2016–2018

   Jukedeck · London, UK 2 papers

   Research at one of the first commercially deployed generative AI music composers. Built deep recurrent neural networks for AI music composition, producing full, well-structured pieces in specified styles and moods. Acquired by TikTok in 2019.

   • Method of generating music data Medeot G., Cherla S., Kosta K., McVicar M., Abdallah S., Selvi M., Newton-Rex E., Webster K. US Patent 11,887,566 · January 2024
   • StructureNet: Inducing Structure in Generated Melodies Medeot G., Cherla S., Kosta K., McVicar M., Abdallah S., Selvi M., Newton-Rex E., Webster K. ISMIR 2018 · blog post
4. 

   PhD Researcher 2012–2016

   City University London · London, UK 11 papers

   Neural probabilistic data modelling for musical sequence analysis using Restricted Boltzmann Machines (RBMs). Developed recurrent temporal discriminative RBMs for learning and generating melodies. Supervised by Prof. Artur Garcez and Dr. Tillman Weyde.

   • Linear-Time Sequence Classification using Restricted Boltzmann Machines Tran S., Cherla S., Garcez A., Weyde T. arXiv 2017
   • Generalising the Discriminative Restricted Boltzmann Machine ICANN 2017
   • Hybrid Dynamic Melody Models ISMIR 2015
   • Discriminative Learning and Inference in the Recurrent Temporal RBM for Melody Modelling Cherla S., Tran S., Garcez A., Weyde T. IJCNN 2015
   • Incremental Dataset Definitions for Large Scale Musicological Research Tidhar D., Benetos E., Wolff D., Dumon E., Cherla S., Weyde T. DLfM Workshop at ACM/IEEE JCDL 2014
   • RNN-based Music Language Models for Improving Automatic Music Transcription ISMIR 2014
   • Multiple Viewpoint Melodic Prediction with Fixed-Context Neural Networks ISMIR 2014
   • A Distributed Model for Multiple Viewpoint Melodic Prediction ⭐ Best Student Paper Cherla S., Weyde T., Garcez A., Pearce M. ISMIR 2013
   • An Efficient Shift-Invariant Model for Polyphonic Music Transcription ML & Music Workshop, ECML 2013
   • A Neural Probabilistic Model for Predicting Melodic Sequences ML & Music Workshop, ECML 2013
   • MIR in Music Education Wiki DMRN+8 Workshop 2013
5. 

   Research Assistant 2011–2012

   Simon Fraser University · Vancouver, Canada 1 paper

   Acoustic modelling research in the TAAP (Transformative Audio and Perception) Lab. Worked on digital waveguide modelling for simulating instrument acoustics, specifically the saxophone.

   • The Saxophone by Model and Measurement SMC 2012
6. 

   MSc, Sound and Music Computing 2010–2011

   Universitat Pompeu Fabra · Barcelona, Spain 2 papers

   Music Technology Group. Master's thesis on automatic melodic phrase continuation using variable-order Markov models, applied to guitar and bass-guitar melodies. Collaborated with PMC Technologies on regression methods for virtual metrology as part of an EU-funded research project.

   • Regression Methods for Virtual Metrology of Layer Thickness in Chemical Vapor Deposition IEEE/ASME Transactions on Mechatronics 2013
   • Automatic Phrase Continuation from Guitar and Bass-guitar Melodies Cherla S., Purwins H., Marchini M. Computer Music Journal 34:3, MIT Press 2013
7. 

   Research Engineer 2007–2010

   Siemens Corporate Technology · Bangalore, India 4 papers

   Research on audio analytics and computer vision. Developed HMM-based systems for continuous environmental audio classification and event detection. Worked on human action recognition in video for surveillance and content indexing applications.

   • Continuous Audio Analytics by HMM and Viterbi Decoding Ramasubramanian V., Karthik R., Thiyagarajan S., Cherla S. ICASSP 2011
   • Audio Analytics by Template Modeling and One-pass DP Based Decoding Cherla S., Ramasubramanian V. INTERSPEECH 2010
   • A Framework for Indexing Human Actions in Video Kulkarni K., Cherla S., Kale A., Ramasubramanian V. ML for Vision-based Motion Analysis Workshop, ECCV 2008
   • Towards Fast, View-Invariant Human Action Recognition Cherla S., Kulkarni K., Kale A., Ramasubramanian V. CVPR 2008 Workshop
8. 

   BTech, Computer Science 2003–2007

   IIIT Hyderabad · Hyderabad, India

   Bachelor's degree in Computer Science and Engineering.

   • Traffic Violation Detection System Undergraduate research project. Built a computer vision system for detecting traffic violations — including red-light running and illegal lane changes — from roadside camera footage, using image processing and machine learning techniques.

Teaching

Teaching assistant at City University London, 2013–2016:

Service

Reviewing — ISMIR (2014, 2015, 2018, 2019) · IJCAI sub-reviewer (2015) · Review Editor, Frontiers in Digital Humanities (Digital Musicology)

Mentoring — Widening Inclusion in Music Information Retrieval (WiMIR) programme, 2018, 2019, 2020, 2025, 2026

Supervision — Sezal Jain (Siemens) · Vaibhav Kumar (City, University of London)

Events — Program Chair, City Informatics PhD Symposium 2014 · Schloss Dagstuhl Seminar on Neural Symbolic Learning · Panelist, Computer Simulation of Musical Creativity Conference