Machine Learning Assistant (Working Student)

About the company

Tools for Humanity (TFH) develops the technology powering World, a global human network built to help people and institutions verify real humans in a privacy-preserving way as AI and autonomous agents become more common online. The product stack includes the Orb for unique human verification, World ID for private proof of personhood, and World App for accessing these capabilities.

World is already operating at a large scale, with more than 17 million verified people across 160 countries and new Orb verifications happening every week. World App is among the most widely used wallets worldwide, and developers are using World ID to create safer online experiences where real people can participate and be recognized.

TFH was founded in 2019 and now has 400+ employees working across hardware, software, AI, cryptography, mobile engineering, and global operations. The company’s teams include people from organizations such as OpenAI, Tesla, SpaceX, Apple, Google, Stripe, Meta, Coinbase, Palantir, and MIT Media Lab. It is backed by investors including a16z, Khosla Ventures, Bain Capital Crypto, Blockchain Capital, Variant, Tiger Global, and Coinbase Ventures.

The company and World have been covered by TIME, Fast Company, Bloomberg, The New York Times, Bankless, and TechCrunch, with leadership also appearing in the Time AI 100.

Role overview

This working-student position is based in Munich (Werksviertel) and is offered through Tools For Humanity GmbH. The role starts immediately and requires a 20-hour workweek, with working days and times arranged flexibly around lectures and exams. This is an on-site position.

You will join the Labelling & Data Collection team, which focuses on building high-quality datasets and internal tools that support robust computer-vision and machine-learning systems. The work is practical, team-based, and well suited to someone who wants hands-on exposure to AI workflows without needing prior ML experience, as training will be provided.

What you will do

• Annotate images and data by adding bounding boxes, labels, and metadata corrections using internal tools and dashboards.
• Check labeled datasets for quality, consistency, and accuracy.
• Assist with dataset preparation for model training, including selection, basic preprocessing, and audit logging.
• Support data-collection activities, including instructions, task validation, and participant metadata where relevant.
• Record issues, help improve annotation guidelines, and contribute to better tooling and workflows, including Streamlit dashboards, MongoDB-backed systems, and internal labelling tools.
• Apply privacy, ethics, and data-handling standards while working with sensitive biometric information.
• Optionally contribute to small Python scripting tasks or join sprint meetings with product and engineering teams.

What we are looking for

• You must currently be enrolled at a university to qualify for the working-student contract.
• You should be very accurate, structured, and comfortable with repetitive tasks that require consistency.
• You need to learn quickly and adapt well to new tools and processes.
• Clear communication is important, including asking questions early, documenting work properly, and collaborating smoothly.
• You should be comfortable using basic technical tools such as web applications and spreadsheets.
• Strong English skills are required; German is helpful but not mandatory.
• Bonus points if you know basic Python, have used databases such as MongoDB or Snowflake, or have previous annotation experience.

Benefits

• Hands-on experience with real ML systems and datasets that directly affect production workflows.
• The chance to suggest improvements to processes and annotation rules, with guidance and review from experienced colleagues.
• A flexible 20-hour schedule that can be arranged around your studies and exam periods.
• A modern office in Munich’s Werksviertel with strong transport connections, refreshments, and daily catering served three times a day.
• Compensation positioned above the market rate for student roles.
• Training in data privacy, ethical handling of biometric data, and internal tooling, plus ongoing mentorship and learning.
• Potential for continued work after graduation or the possibility of completing a thesis project with the team.

Application details

Applicants are asked to send a CV and a short cover letter of one to two paragraphs. The message should include the earliest possible start date, weekly availability in hours, and the current degree program plus year of study (matriculation).

Location

Munich, Bavaria, Germany, specifically Werksviertel.

Additional information

Company: Tools For Humanity GmbH. Contact person: Constantin Ingelheim. Address: August-Everding-Straße 25, 81671 Munich, Germany. Email: constantin.ingelheim@toolsforhumanity.com.

This role is part of a working-student setup with flexible scheduling and is intended for candidates who can work 20 hours per week.

No prior machine-learning experience is required.