CEO / Project Manager / Developer at Trigonella GmbH (self-employed). I build practical machine-learning and NLP systems, ranging from production-grade data pipelines and RAG assistants to domain-specific solutions in education, finance, healthcare imaging, and media intelligence. I care about requirements, documentation, transparency, and rapid delivery of business value. My work combines solid software engineering (Python/PHP/SQL/CI/CD) with modern ML/LLM tooling (PyTorch/TensorFlow, transformers, LangChain/RAG), and I enjoy turning complex constraints into usable products for real stakeholders.

Dominik Heeb

CEO / Project Manager / Developer at Trigonella GmbH (self-employed). I build practical machine-learning and NLP systems, ranging from production-grade data pipelines and RAG assistants to domain-specific solutions in education, finance, healthcare imaging, and media intelligence. I care about requirements, documentation, transparency, and rapid delivery of business value. My work combines solid software engineering (Python/PHP/SQL/CI/CD) with modern ML/LLM tooling (PyTorch/TensorFlow, transformers, LangChain/RAG), and I enjoy turning complex constraints into usable products for real stakeholders.

Available to hire

CEO / Project Manager / Developer at Trigonella GmbH (self-employed). I build practical machine-learning and NLP systems, ranging from production-grade data pipelines and RAG assistants to domain-specific solutions in education, finance, healthcare imaging, and media intelligence.

I care about requirements, documentation, transparency, and rapid delivery of business value. My work combines solid software engineering (Python/PHP/SQL/CI/CD) with modern ML/LLM tooling (PyTorch/TensorFlow, transformers, LangChain/RAG), and I enjoy turning complex constraints into usable products for real stakeholders.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Intermediate
See more

Language

English
Fluent
German
Fluent
French
Advanced
Italian
Intermediate
Spanish; Castilian
Intermediate
Swedish
Beginner
Romansh
Advanced

Work Experience

Student AI/ML at EPFL Extension School
April 1, 2022 - August 27, 2025
Utilized ridge regression for real estate price prediction with 9.0% MAE and conducted binary sentiment analysis using bag-of-words techniques.
Extension school AI/ML Student (remote) at EPFL Extension school
February 1, 2021 - April 30, 2022
AI/ML training and applied projects. Built datasets and trained models for computer vision tasks (e.g., identifying build years of locomotives from images) and vehicle classification; also performed sentiment analysis using bag-of-words.
CEO / Project Manager / Developer at Trigonella GmbH
October 1, 2001 - Present
Self-employed since 2001. Leads development across machine learning, NLP/LLMs, software engineering, and applied analytics. Delivered systems such as an intelligent childcare spot search platform (ranking/matching and data synchronization), multilingual media classification and forecasting (RF/XGBoost/neural/transformers + ARIMA/SARIMA/Prophet/LSTM), and scenario-based investment advisory document generation by preparing data from mainframe to Microsoft Office outputs. Also produces numerous public ML/NLP applications and research-grade prototypes in domains including document digitization, computer vision, knowledge graphs/ontologies (RDF/SPARQL), and minority-language NLP.
Developer at Trigonella GmbH
October 1, 2001 - Present
Developed and deployed a binary computer vision classifier to distinguish between intact and damaged chocolate pralines on a conveyor belt using XGBoost and PCA, achieving 86% accuracy. Successfully deployed with Flask, SSH, and Bash on DigitalOcean. Implemented a CI/CD pipeline with GitHub Actions, pytest, and flake8 to maintain code quality and PEP8 compliance. Created a RAG chatbot about Swiss artist Ruth von Fischer using LangChain and OpenAI, accessible publicly. Expanded text classification frameworks for commsLAB, targeting multilingual corpora with Spacy and transformer models. Enhanced NLP pipelines with semi-supervised learning, feature engineering, and machine learning. Delivered machine learning models for company valuations (Nimbo AG), job fitness assessments, customer targeting, and satisfaction prediction. Developed multilingual chatbots and performed audio transcription and analysis, as well as morphological generation and spell-checking for Vallader (Romansh).
Software Developer at e-Serve
August 31, 2000 - August 27, 2025
Contributed to the development of a rule-based chatbot for customer inquiries at Coop. Developed morphological generators for German and French languages to enhance linguistic processing for chatbot applications.
Software Engineer at Sirius AG
October 1, 1998 - September 30, 2001
VBA
School teacher and programmer at Gymnasium Muttenz
August 1, 1990 - June 30, 1992
Taught programming and developed simple educational software, including a Pascal program for hand-written digit detection using a single-layer perceptron with semi-supervised learning.

Education

Bachelor of Science at University of Basel
October 1, 1983 - May 15, 1986
Biology
Master of Science at University of Basel
May 15, 1986 - August 31, 1989
Biology, Mathematics
COS: "Data Science and Machine Learning" at ETH Lausanne (EPFL)
February 15, 2021 - April 15, 2022
Python stack: Shallow Learning, Deep Learning, NLP, Computer Vision
MSc Natural Scientist (Evolutionary Biology, Math, IT, Stats, Physics, Chemistry) at Basel University
January 11, 2030 - August 27, 2025
Applied Data Science / Machine Learning at ETH Lausanne (EPFL)
January 11, 2030 - August 27, 2025
AI Industry-Readiness Residency Program at Apziva, Ankara
January 11, 2030 - August 27, 2025
MSc Natural Scientist (Evolutionary Biology, Math, IT, Stats, Physics, Chemistry) at University of Basel
January 11, 2030 - August 9, 2026
Applied Data Science / Machine Learning (AI Industry-Readiness Residency Program, Apziva, Ankara) at ETH Lausanne (EPFL) - AI Industry-Readiness Residency Program (Apziva, Ankara)
February 1, 2021 - April 30, 2022

Qualifications

Hands-on course on Big Data, Apache Spark, Hadoop, and Kubernetes
January 11, 2030 - August 27, 2025

Industry Experience

Agriculture & Mining, Financial Services, Professional Services, Education, Media & Entertainment, Software & Internet, Healthcare, Government
    Estimating house prices
    Applying various regression techniques to estimate the value of a house on sale. Thorough EDA efforts.
    Forecasting
    Forecasting the rentals at a bike shop depending on weather conditions and weekday information.
    Estimating equity values
    Using various regression techniques to estimate the equity values of a company. The input data consists both of hard business indicators as well as soft survey information on the company.
    BLEU scores
    Evaluating NLP output using the BLEU score metric.
    Natural Language Processing on media articles.
    Classifying a corpus of 1.6 million media articles by a set of given classes.
    RAG
    Creating a chatbot using LangChain RAG techniques: Retrieval augmented generation.
    Web scraping
    Data collection using web scraping techniques.
    Image segmentation
    Image segmentation: SAM and YOLO
    Scoring models and recommender models
    Smart database filters using scoring techniques.
    Computer Vision AI
    Image classification and OCR