Data Scientist

Building AI solutions on the Edge.

I am a PhD candidate at Loughborough University based in the UK. My PhD research topic is 'Computationally Efficient Embedded Classification Algorithms for Augmentative and Alternative Communication (AAC)'. I work on researching efficient AI solutions for assistive technologies that help individuals who are speech disabled. Partnering with me is the Ace Centre, a charity that deals with assistive technologies and AAC. I hold a master's degree (with distinction) in Data Science awarded at Loughborough University, as well as a bachelor's degree (first class honours) in Electrical and Electronic Engineering. My passion lies in working with artificial intelligence to make a positive impact in society.

View Work Download CV
K. Simatwo

Technical Arsenal

AI & Signal Processing

Python, MATLAB, PyTorch, Scikit-learn, Transformers (BERT), Wavelet Analysis, Statistical Modeling

Edge & Hardware

Raspberry Pi, Arduino, STM32, Real-time Systems, Edge Algorithms

Software Engineering

Git, SQL, Jupyter, Algorithm Optimization

Experience

Jan 2023 - Present

PhD Researcher

Loughborough University

  • Architecting a computationally efficient, real-time breath pattern recognition system for assistive speech on Raspberry Pi.
  • Benchmarked high-performance classification models against 1NN-DTW baselines.
  • Contrasted Deep Learning (GoogLeNet on CWT) vs. Classical Wavelet approaches.
  • Achieved sub-second prediction times for genuine real-time communication.
  • Mentored undergraduate students on STM32 hardware integration.
Jan 2024 - Jul 2024

Data Science and AI Intern

Turing Internship Network

  • Developed a semantic similarity metric using BERT embeddings as an alternative to Word Error Rate.
  • Validated non-reference audio quality metrics adopted by partner teams.
  • Applied unsupervised clustering to identify acoustic patterns in speech corpora.
Jun 2021 - Aug 2021

Research Intern

Insigneo Institute for in silico Medicine

  • Engineered a control system for a flow phantom integrating encoders, Doppler probes, and lasers.
  • Developed a high-precision laser velocity sensor achieving 98% measurement accuracy.
  • Optimized MATLAB analysis code and designed 3D-printed casings for user interfaces.

Projects & Initiatives

Tiger Stripe Classification

The Alan Turing Institute

Collaborated with the Environmental Investigation Agency to classify tigers. Executed EDA on imbalanced datasets and validated a PyTorch pipeline with masking to enhance accuracy.

Microsoft Embrace Hackathon

1st Place Winner

Led a 4-person team to deploy a cloud-based sentiment analysis app on Azure. Engineered real-time Twitter data analysis for hate speech detection.

Enactus Sheffield

Innovation and Technology

Technical consultant for student-led projects. Designed the official website and created sustainability checklists for commercial portfolios.

Education

PhD (AI in Assistive Speech)

Loughborough University | 2023 - 2026

MSc Data Science

Loughborough University | Distinction

BEng Electrical & Electronic Engineering

University of Sheffield | First Class Honours

Selected Publications & Talks