Programming & Data Stack
Core language and framework foundations for building analytical systems and production ML services.
- Python
- SQL
- Pandas / NumPy
- Plotly
ML engineer building for real-world pressure
Ronald Piku.
ML Engineer & Data Scientist
I design machine learning and analytics systems that help people make better decisions when speed, trust, and operational reliability all matter. My work spans Formula 1, applied ML, and product-focused data tooling.
Skills & Expertise
A toolkit forged in high-pressure environments where speed, accuracy, and trust are non-negotiable.
Machine Learning & AI
Applied ML skills across the full lifecycle, from feature design to reliable production delivery.
- Computer Vision
- LLMs
- Statistical Modelling
- ML Libraries (scikit-learn, PyTorch)
- Feature Engineering
- Model Evaluation & Validation
Data & Analytics
Decision-focused analytics and data foundations that scale from exploration to production.
- Data Analysis & Visualisation
- Data Pipelines
- Performance Reporting
- Database Design & Query Optimization
DevOps & Tools
MLOps and engineering practices that keep systems deployable, observable, and resilient.
- Docker
- CI/CD (GitHub Actions)
- Git
- Agentic Workflows
- Model Deployment & Serving
How I got here
From markets to machine learning systems built for real-world pressure
Alpine Formula One Team
Vehicle Performance Software Engineer
In Formula 1, every decision sits on a clock. I build performance tooling that helps engineers and analysts move from raw telemetry to confident calls before, during, and after race weekends.
Core focus: Real-time analytics, race engineering decision support, and reliable reporting under tight operational deadlines.
Built with
Python
Plotly
SQL
Vehicle Dynamics
Data Analysis
MATLAB
Turning points
- Built agentic workflows to aid race engineers in quickly diagnosing performance issues and making informed setup decisions.
- Built internal tooling that reduced analyst turnaround time and improved race-week reporting reliability.
- Delivered pre-race and post-race analysis workflows that improved setup trend visibility and supported faster trackside decisions.
- Replaced a legacy statistic system with a faster Python pipeline, lowering operational cost and maintenance overhead.
- Built CI/CD pipelines and monitoring for critical data systems, improving reliability and reducing downtime during race weekends.
- Provided event support for live data systems, ensuring smooth operation and quick issue resolution under high-pressure conditions.
Behind the scenes +
Much of the value came from reducing cognitive load under pressure. Fewer manual checks, clearer visual cues, and faster handoffs between analysis and action.
Pineamite
Machine Learning Engineer
This was where I learned to treat ML as a product, not just a model. I built production-oriented computer vision systems that made live sports coverage faster, more cost-effective, and more dependable.
Core focus: Computer vision model development, data pipeline design, and performance monitoring for live environments.
Built with
Machine Learning
Data Engineering
Python
Computer Vision
PyTorch
ONNX
Cloudflare
Turning points
- Led a team of engineers to release a native application within 12 weeks through structured planning and cross-functional execution.
- Built a computer vision model that improved the accuracy and reliability of live sports coverage, enhancing viewer experience while reducing operational overhead.
- Designed end-to-end data pipelines that reduced live event coverage cost while maintaining reliable output quality.
- Created and curated a proprietary training dataset that improved benchmarking and model iteration speed.
- Introduced monitoring and performance checks that surfaced bottlenecks earlier and reduced processing latency.
Behind the scenes +
The hardest part was balancing quality and latency simultaneously. Better data curation and observability often delivered bigger wins than model tweaks alone.
Summer Boarding Courses
Engineering Teacher
Teaching sharpened one of my strongest engineering skills: making complexity usable. I designed and delivered learning experiences that turned abstract concepts into practical outcomes.
Core focus: Technical communication, curriculum design, and structured problem solving with diverse learner groups.
Built with
Teaching
Curriculum Design
STEM
Mentorship
Problem Solving
Turning points
- Built and delivered an engineering curriculum for students aged 13-18, balancing rigor with hands-on creativity.
- Guided learners through practical projects from self-propelled systems to bridge design and game development.
- Strengthened the ability to explain difficult concepts in a way teams can execute quickly.
Behind the scenes +
This role improved how I structure technical communication by starting from intent, reducing jargon, and defining the next concrete action.
Personal Development
Career Break
I used this period as a deliberate reset rather than a pause. I invested in ML foundations while teaching abroad and broadening my problem-solving perspective.
Core focus: Structured upskilling in machine learning and data science, with deliberate practice on applied projects.
Built with
Learning
Travel
Machine Learning
Data Science
Turning points
- Completed focused ML and data science study to build stronger mathematical and implementation fundamentals.
- Applied learning through practical work and teaching, improving adaptability and communication under changing conditions.
Behind the scenes +
This chapter strengthened consistency and self-direction, habits that later made it easier to ship under pressure in industry roles.
Veblen Collectables
Commodities Broker
My early career was in markets where uncertainty was constant and decisions were public. I worked in high-pressure trading environments where research quality, timing, and trust directly shaped outcomes.
Core focus: Decision-making under uncertainty, stakeholder management, and disciplined market analysis.
Built with
Research
Negotiation
Client Strategy
Decision Making
Market Analysis
Turning points
- Managed a portfolio of high-value assets with disciplined research and strategy-led decision making.
- Built strong client trust through clear communication and consistent delivery in volatile conditions.
- Developed calm, structured execution habits that transfer well to production engineering and ML operations.
Behind the scenes +
This experience shaped my approach to engineering trade-offs: evaluate risk, communicate clearly, and commit decisively when timing matters.
Education
Academic foundation in Machine Learning and Mechanical Engineering
MSc AI for Engineering and Design
University of Bath
Oct 2024 - Oct 2026
Expected: Distinction
BEng Mechanical Engineering
University of Leeds
Oct 2018 - Oct 2021
Classification: 2:1
Featured Projects
Stories of building practical ML products and data systems, with visual proof and hands-on paths to explore each project.
Production-style build
MatchMind
From event data to match-winning insight
Coaches need answers quickly, not more dashboards. MatchMind compresses hours of manual analysis into fast, decision-ready outputs.
Impact: Built to reduce analysis turnaround from hours to minutes for repeatable match workflows.
Public repository
Python
SQL
Data Pipelines
Plotly Dash
Performance Analysis
Research project
Towards Simulation-to-Reality Transfer in EIT Tactile Sensing
Teaching models to handle real-world sensor noise
Real sensors are messy. This work explores how noise-aware training can make tactile ML systems more robust before they ever touch live hardware.
Impact: Reduced dependence on expensive real-world data collection while improving transfer robustness.
Public repository
Python
Machine Learning
Deep Learning
Robotics
Sim2Real
Private product
AlphaScreen
AI-assisted equity research workflow
AlphaScreen is designed for fast conviction: less time wrestling fragmented data, more time validating investment hypotheses.
Impact: Built to make multi-factor equity research workflows faster, clearer, and easier to repeat.
Private codebase
Agentic Workflows
Python
NLP
LLMs
FinTech