Book a call

Pick a time that suits you. You'll get a confirmation email with a calendar invite.

Island map

The journey

Places

Nataell Cornu

Start

Who I am

Software engineer and data scientist in Zürich.

I use AI to free people from monotonous work. I understand problems from the ground up and independently build solutions that are faster and cheaper than usual.

Personality

Independent: I decide and shape things.
Proactive: I build solutions, often before anyone asks.
Team player: One goal, everyone owns their part.
Analytical: Understand first, then build.

Email me LinkedIn

Hobbies

Then and now

Sport has been part of my life since I was a child: football and table tennis back then, running, cycling and the gym today.

Then
- Football
- Table tennis
Today
- Running
- Cycling
- Gym

Education

2017–2025

MSc Computer Science
University of Zurich · 09.2023 – 06.2025
- Major: Software Systems
MA Business and Economics
University of Zurich · 09.2020 – 07.2023
- Major: Management and Economics, minor: Economics
- Built interactive R Shiny applications for data analysis
BA Business and Economics
University of Zurich · 09.2017 – 07.2020
- Major: Economics, minor: General Economics

Experience

2019–2026

Software Engineer & Data Scientist
Acosom · Freelance consulting · 06.2025 – 06.2026
- On-premise serving of AI and LLM models in production
- Multi-GPU inference and deployment with vLLM, SGLang and TensorRT
- Served models over real-time data streams with Apache Flink
- Data and software engineering on client consulting projects
Software Engineer & Data Scientist
Meta-Sys, Bubikon · Part-time · 06.2019 – 06.2025
- Software development and data management
- Designed and maintained ETL pipelines
- Data parsing with multimodal LLMs and retrieval-augmented generation
- Data cleaning, feature engineering and geospatial data analysis
- Machine learning for inference and clustering: XGBoost, ensembles, DBSCAN, k-means
- Statistical modelling in R: price indices, price estimation and causal analysis with OLS, GLMs, GAMs and regularised regression
- Dynamic data visualisations in R with ggplot2, Shiny and base plots

Projects

Today

Motivation

I used to extract data from documents by hand. I know how monotonous it is and how much time gets lost. Today I use AI to build the solutions that take over exactly this work.

Value I deliver

95 % less time spent
MiruIQ, co-founder
A client spent three weeks a month entering data by hand. Today he only reviews the result.
OCR for millions of documents
TurboOCR, author
Faster and cheaper text recognition than standard tools, available as open source.
AI kept in-house
On-premise, in production
LLMs run on own infrastructure, with multi-GPU inference and real-time data streams.

Skills

Toolbox

Key competencies

AI, LLMs and RAG in production
Automated data extraction and OCR
Software engineering
Data science and statistics

Skills

Python
R
rpy2
SQL
Java
C
ReactJS
LLMs
RAG
OCR
NLP
vLLM
On-prem serving
Apache Flink
Apache Airflow
ETL
Docker
Causal inference
Econometrics
DevOps

Programming languages

Python: 9 years
R: 6 years
JavaScript: 5 years
Java: 4 years
MATLAB: 2 years
C++: 1 year

Languages

German: Native
French: Native
English: Professional
Italian: Basic

Contact

What's next

Book a call straight into my calendar, or write me an email.

Where I make an impact

In teams with a lot of manual data and document work that they want to automate with AI.

Email me LinkedIn

Zürich, Switzerland