About Me
Hi there! I'm Toni.
"I see research, engineering, and IT as the bridge between cutting-edge science and real-world impact."
I'm a process engineer who builds the data and AI systems for the fab floor. At Murata I run the silicon–glass fusion area of a wafer fab: qualifying equipment, leading development projects into production, and chasing yield. The same job produces the data, so I also build the pipelines, models, and automation that turn that data back into process decisions. One loop, both ends.
I hold a Master of Science from Aalto University (Chemical, Biochemical and Materials Sciences), which trained me to work from first principles and experimental validation, and I am completing a Bachelor's Degree in Data and AI at KAMK alongside full-time work (planned graduation: 2027). I have worked in semiconductor manufacturing since 2012, starting as a cleanroom operator at Okmetic during my studies. I know what a bad shift on the tool looks like from the operator's side, which is why I care so much about automation that prevents mistakes.
Outside work I ride my mountain bike through forest trails, build my digital homelab, and connect the digital world to the real one with 3D-printed models. I'm also a big fan of science fiction, which inspired the green-on-black aesthetic of this site.
Mission
{
"name": "Toni Kiuru",
"role": "Process Engineer x AI/Data Engineer (Real World <-> Data <-> AI/ML)",
"core_competencies": [
"Process Ownership & Statistical Qualification",
"Predictive Systems & Production ML",
"Cross-functional Team Lead"
],
"mission_vector": "Turning Real-World Data into Scalable, Intelligent Systems",
"status": "In Progress"
}
Example: From Raw Signals to AI-Driven Impact
Here is one concrete example of how I work, drawn from semiconductor manufacturing.
It starts with people, not data. Before I open a dataset, I interview the experts: the operators, the technicians, the engineers who have lived with the process. I want to understand the current situation as they see it, where the real pain is, and what questions actually need answering. Only then do I turn to the data and ask the harder questions: does this data describe reality, or just what the tool happened to log? Does it answer the process challenge we have, or a different one? This approach, combined with my process engineering experience, allows me to see the unknowns, understand the process well enough to know what should be measurable, and pinpoint the data we are missing before anyone builds a model on top of the gap.
Then the modeling. I design models that capture temporal evolution across full process runs, giving accurate and physically consistent representations of system behavior. My modeling philosophy centers on rigorous validation: structured dataset partitioning, robustness analysis, and continuous monitoring of generalization under drift across tools, materials, and production conditions.
This validation discipline comes straight from process engineering. On the fab floor, nothing changes production until it survives a statistical qualification: a designed experiment, a capability study, a measurement system that has been proven capable of seeing the effect at all, and acceptance criteria written down in advance. A model has to clear the same bar as a new tool, a recipe change, or a supplier before it gets anywhere near production.
The domain here happens to be manufacturing, but the pipeline itself is universal: any field with messy real-world data (logistics, energy, healthcare, finance) follows the same path from raw signals to models to decisions. That end-to-end path is exactly where I want to work.
How raw signals become business impact:
What I Bring to the Team
Innovation starts with bringing people together. Through active listening, I help teams formulate actionable steps toward transformative change. I specialize in structuring messy data, building machine learning models that work in production, and connecting data insights with real physical understanding, such as predicting material behavior from raw process data.
I enjoy shaping ideas into clear project plans, coordinating between R&D, IT, and production, and communicating results clearly to researchers, operators, and management alike. I lead cross-functional teams through root-cause investigations when a critical process breaks, and I lead projects the other way too — taking an idea through experiments, analysis, documentation, and formal approval until it is running in production. I am a native Finnish speaker and work fluently in English across multicultural environments.