Portrait of Rui Su
AI Engineer · PhD Researcher · Founder

Rui Su

Machine Learning Engineer and PhD candidate at Maynooth University, specialising in real-time ML systems, ensemble prediction, and low-latency algorithm design. Founder & Director of Intonation Labs (Singapore), building production AI systems and SaaS products.

Singapore / Dublin, Ireland PhD completion: Dec 2026 Open to work

I build production AI systems that work under real-world constraints — latency budgets, cost limits, and reliability requirements. My PhD research at Maynooth University focuses on real-time audio processing and adaptive tuning algorithms, combining self-supervised ML, ensemble prediction, and low-latency signal processing.

In February 2026, I founded Intonation Labs Pte. Ltd. in Singapore, operating a dual-engine model: B2B ML consulting for startups and SMEs, and development of SaaS products including IntonationAI (AI music coaching), IntoMeeting (real-time AI meeting co-pilot), IntoMarketing (AI marketing automation), and IntoBadminton (evidence-led badminton equipment finder).

My path from a B.A. in Psychology at UC Davis to a Computer Science PhD gives me a strong appreciation for human-centred design in technical systems. I am bilingual in English and Mandarin Chinese, and have lived and worked across three continents.

Founder & Director

Feb 2026 – Present
Intonation Labs Pte. Ltd. · Singapore
  • Founded an AI consulting and product company operating a dual-engine model: B2B ML consulting for startups/SMEs and development of four SaaS products
  • Built and deployed IntonationAI — AI-powered vocal coaching platform using Vertex AI Gemini 2.0 Flash, Google Speech-to-Text v2, and Firebase
  • Built and deployed IntoMeeting — real-time AI meeting co-pilot with WebSocket streaming, RAG-powered knowledge base search, and live question detection
  • Architected production systems on GCP (Cloud Run, Firestore, Cloud Functions) with Stripe billing and Terraform IaC

Technology Engineer

Sep 2025 – Present
Ubimus Conference · Remote
  • Designing and deploying the conference web platform for an international Ubiquitous Music research symposium
  • Building responsive web interfaces using modern JavaScript and web technologies

Firmware Test Engineer

Jun 2022 – Dec 2023
Xunison · Dublin, Ireland
  • Tested and validated 5G router firmware for IoT home automation platform
  • Performed command-line debugging and configuration to optimise network performance

PhD, Computer Science

2023 – Dec 2026
Maynooth University, Ireland

Thesis: Instant Harmonies — Enhancing Music with Web-Based Just Intonation Tuning Algorithms

Research: Low-latency audio processing, self-supervised key detection, ensemble ML prediction, production deployment. Supervisors: Prof. Joseph Timoney, Prof. Damián Keller.

Higher Diploma, Software Development

2021 – 2022
Maynooth University, Ireland

First Class Honours. Data Structures & Algorithms, Database Systems, Web Development, Software Engineering.

M.A., Child, Youth and Family Studies

2020 – 2021
Institute of Technology Carlow, Ireland

NFQ Level 9. Applied psychology, research methods and dissertation, and leadership, strategy & governance.

B.A., Psychology (Minor: Economics)

2015 – 2020
University of California, Davis

Research methodology, statistical analysis, and experimental design.

Machine Learning & AI

PyTorch TensorFlow scikit-learn Hugging Face Transformers RAG LangChain LangGraph Agentic AI Prompt Engineering Ensemble Methods Real-Time Inference

Full-Stack & Cloud

Python JavaScript / TypeScript Next.js React FastAPI Flask GCP Firebase Cloud Run Vertex AI Terraform Docker

Audio & Signal Processing

WebAudio API WebMIDI API DSP Audio Fingerprinting Low-Latency Streaming MIDI Protocol

Data & Infrastructure

NumPy Pandas SQL Firestore WebSocket REST APIs Git Linux Stripe API

Instant Harmonies

PhD Research

Production-grade browser-native system for real-time music analysis and adaptive just intonation tuning. Combines ensemble ML prediction with low-latency audio processing (~1.7 ms median / ~2.2 ms p99 end-to-end latency).

Three research contributions: (1) S-KEY-Symbolic — first self-supervised local key detection for symbolic MIDI, (2) MTS/MPE dual-protocol real-time tuning with hierarchical fallback, (3) BACHI chord-aware JI tuning with 7-limit ratios from 99,251 extracted chord labels.

PyTorchFlaskWebMIDIWebAudioJavaScriptParangonar
Live demo GitHub

IntonationAI

Live

AI-powered vocal coaching platform providing personalised, real-time feedback on pitch accuracy, tone, timing, and technique. Uses Vertex AI Gemini 2.0 Flash and Google Speech-to-Text v2 for audio analysis.

Next.jsFastAPIVertex AISpeech-to-TextFirebaseStripe
intonationai.app

IntoMeeting

Live

Real-time AI meeting co-pilot that runs alongside Google Meet or Zoom. Captures meeting audio via streaming transcription, detects client questions, and surfaces RAG-powered answers from the user's knowledge base in real time.

FastAPIWebSocketGeminiRAGFirestoreNext.jsStripe
intomeeting.app

IntoBadminton

Live

Evidence-led badminton equipment finder and review site. A five-step finder captures skill level, discipline, and play style, then returns a ranked shortlist with transparent fit-score reasoning — backed by structured comparison guides with explicit source-authority labelling.

Next.jsReactTypeScriptTailwind CSSGitHub Pages
intobadminton.com

IntoMarketing

In Development

AI-powered marketing automation platform with daily intelligence pipelines, automated post generation, lead qualification, and scheduled social publishing to LinkedIn and X.

FastAPICloud FunctionsVertex AITerraformNext.jsStripe

A Multi-Protocol Adaptive Microtuning Architecture: Bridging MTS, MPE, and MIDI 2.0 for Struck-String Just Intonation

Sound and Music Computing Conference (SMC 2026) Published
R. Su, J. Timoney, D. Keller
Download PDF

A Multi-Modal Adaptive Microtuning Architecture: Bridging MTS, MPE, and MIDI 2.0 for Real-Time Just Intonation in Ubiquitous Music

GIM 2026 — International Graduate Residency in Ubiquitous Music Published
R. Su, J. Timoney, D. Keller
Download PDF

A Multi-Modal Adaptive Microtuning Architecture for Real-Time Just Intonation on the Web

3rd Symposium on Digital Art in Ireland, UCC & Sample-Studios · Apr 2026 Published
R. Su, J. Timoney, D. Keller
Download PDF (symposium booklet)

MIDI Adaptive Tuning Strategies by Means of AI-Based Struck-String Interaction in Ubimus

Ubimus Journal · Aug 2025 Published
R. Su, J. Timoney, D. Keller
Download PDF

Just Intonation and Inharmonicity in Struck-String Interaction

Ubimus Journal · Oct 2024 Published
R. Su, J. Timoney, D. Keller
Download PDF

Professional certificates and specialisations completed via Coursera. Click any certificate to view or download the PDF, each of which includes a Coursera verification link.

Imperial College London

Let's build something together.

I am open to ML engineering roles, AI consulting engagements, and research collaborations. Based in Singapore, open to on-site, hybrid, or remote.