Co-Founder & Head of Science — ByteFuse AI
Ruan van der Merwe
I’m an applied machine learning researcher and AI leader based in Cape Town, working on speech and language systems with a focus on continual and meta-learning for acoustic modelling. My research addresses catastrophic forgetting in few-shot speech classification, analyses representation geometry to predict downstream task performance, and develops principled evaluation frameworks for real-world speech systems.
I co-founded ByteFuse AI, where I lead scientific strategy and ML architecture — guiding research direction, mentoring researchers, and translating algorithmic advances into scalable deployed systems across education, speech, and analytics.
Find me on Google Scholar, GitHub, LinkedIn, or by email at ruan [at] bytefuse [dot] ai.
Recent writing
all posts →Selected papers
all papers →- 2026Scaling Few-Shot Spoken Word Classification With Generative Meta-Continual LearningarXiv preprint arXiv:2605.13075
- 2026Does Language Matter for Spoken Word Classification? A Multilingual Generative Meta-Learning ApproacharXiv preprint arXiv:2605.13084
- 2023Mitigating Catastrophic Forgetting for Few-Shot Spoken Word Classification Through Meta-LearningInterspeech 2023, pp. 441–445
- 2022Manifold Characteristics That Predict Downstream Task PerformanceFirst Workshop on Pre-training: Perspectives, Pitfalls, and Paths Forward at ICML 2022
Selected projects
all projects →- 2021 — presentByteFuse AIAI R&D company backed by a R55M investment from Novus Holdings, building enterprise-scale AI across education, speech, and analytics.
- 2023 — presentMaskiAI-powered WhatsApp tutoring companion built with Maskew Miller Learning, serving South African learners with CAPS-aligned, personalised support.
- 2023MAMLCon (open source)Reference implementation for Mitigating Catastrophic Forgetting for Few-Shot Spoken Word Classification Through Meta-Learning (Interspeech 2023).