ML Internship: Deep Learning for Causal Inference

Ethonai in Zürich offers an internship in deep learning for causal inference, focusing on industrial data. Candidates will explore ML methods to uncover causal relationships in manufacturing data, working on prototypes with mentorship from ML experts.

This internship sits at the crossroads of deep learning and industrial applications, aiming to use modern ML techniques to infer causal structures from factory data. The goal is to develop working prototypes for Root Cause Analysis, turning innovative ideas into practical tools for the analytics stack.

Responsibilities

• Research and evaluate deep learning methods for causal inference on tabular data, identifying the most viable approaches for industrial use.
• Create models that combine tabular data with natural language context to enhance causal discovery and root cause analysis.
• Develop infrastructure for experiments and benchmarking, comparing prototypes against existing causal discovery methods using public and anonymized manufacturing data.
• Collaborate with ML scientists and engineers to evolve prototypes into components that could integrate into the analytics platform.

Requirements

• Currently enrolled in or recently completed a BSc or MSc in Computer Science, Data Science, Machine Learning, Math, Statistics, or a related field.
• Strong Python programming skills, including experience building training pipelines, simulations, and benchmarking systems from scratch.
• Solid understanding of machine learning fundamentals and hands-on experience with deep learning frameworks like PyTorch or JAX beyond academic coursework.
• Excellent verbal and written communication in English.
• Availability for at least 6 months, with potential for longer duration.
• High motivation to work in a fast-paced environment and drive ideas from concept to functional prototype.

Nice to have

• Experience with deep learning models for structured data, such as TabPFN, TabICL, FT-Transformer, or TabNet.
• Background in causal inference methods, including PC, GES, NOTEARS, LiNGAM, DoWhy, or score-based and diffusion-based approaches.
• Familiarity with graph neural networks or structured prediction techniques.
• Experience integrating textual or semantic data with numerical inputs, such as using LLM embeddings or multimodal models.
• Knowledge of classical statistical modeling, Bayesian inference, or probabilistic programming.
• Experience with time-series data from industrial or physical systems.
• Demonstrable open-source contributions or personal projects showcasing strong coding skills.

What the company offers

• Direct mentorship from ML scientists and the engineering team.
• An ambitious, well-defined project with room for personal contribution and innovation.
• On-site work in a modern office in central Zürich, with flexible scheduling options.

About the company

Ethonai is a leader in Industrial AI, focused on reducing waste in manufacturing. The company serves major global manufacturers and draws talent from top tech firms and consulting firms. It is backed by leading investors and fosters a culture of curiosity and mutual growth.

• Committed to solving the trillion-dollar waste problem in manufacturing.
• Serves top global manufacturers including Siemens, Bosch, Lindt, and Roche.
• Team members come from companies like Google, Meta, Palantir, and MBB.
• Backed by top investors such as Index Ventures, General Catalyst, and Earlybird.
• Values curious, growth-oriented individuals.
• Supports professional and personal development for all team members.