Data & AI engineer for ecosystem tools

JUNI; Berlin; Hybrid; EUR 70,000 - 110,000; Full time

Benefits offered by this job

Flexible working hours

Hybrid work

Berlin office

Job summary

JUNI приглашает инженера по данным и искусственному интеллекту в Берлине. Вы возьмете на себя долгосрочную ответственность за JUNI Core: пайплайны, модель данных и методы ИИ, позволяющие данным быть полезными.

Плюс — объединять источники, обеспечивать качество и согласованность данных, работать с командой продукта и техническими специалистами. Вы будете подключать внешние источники, применять эмбеддинги и LLMs для сопоставления, классификации и обогащения данных, документирования и тестирования

Qualifications

• Опыт построения и эксплуатации пайплайнов обработки данных в продакшн.
• Уверенное владение SQL и PostgreSQL.
• Знание Python для работы с данными (pandas, ноутбуки, скрипты оценки).

Responsibilities

• Take responsibility for JUNI Core: connecting all source systems, the data model and operations
• Build and operate pipelines between our tools and the platform in both directions
• Keep data reliable: a clear source for every field, syncs without duplicates, and alerts when something goes down
• Connect new sources and decide how they fit into the model
• Work with the Product & Tech team to determine what data JUNI needs, where it comes from and what we are allowed to use it for
• Build the data foundation for new tools in our ecosystem work
• Integrate external data such as publications, patents or company data via APIs and scraping, and link it to our existing data
• Ensure that users can see where a result comes from
• Use LLMs and embeddings for matching, classification and enrichment, with evaluation built in
• Define what good results look like for each use case: acceptable error rates, reference cases and what happens if a result is wrong
• Work with the Technical Product Manager and subject-matter leads to determine what is worth automating
• Documentation and tests are part of the work, not something done afterwards

Skills

SQL

PostgreSQL

Python data work

APIs & webhooks

LLMs in pipelines

Matching methods

English fluent

Record linkage

Tools

Pandas

Notebooks

Evaluation scripts

Claude Code

Web scraping

Embeddings

Job description

Описание

JUNI is the startup factory for science-based spin-offs in Berlin and Brandenburg. We connect universities, startups, companies, investors and policymakers, and create the spaces, programs and partnerships that turn research into successful companies. With JUNI Core, we are building a central data platform that brings everything together and makes it available to our teams’ tools. As a Data & AI Engineer, you will take long-term responsibility for this data layer: the pipelines, data model and AI methods that make the data usable.

Задачи

• Take responsibility for JUNI Core: connecting all source systems, the data model and operations
• Build and operate pipelines between our tools and the platform in both directions
• Keep data reliable: a clear source for every field, syncs without duplicates, and alerts when something goes down
• Connect new sources and decide how they fit into the model
• Work with the Product & Tech team to determine what data JUNI needs, where it comes from and what we are allowed to use it for
• Build the data foundation for new tools in our ecosystem work
• Integrate external data such as publications, patents or company data via APIs and scraping, and link it to our existing data
• Ensure that users can see where a result comes from
• Use LLMs and embeddings for matching, classification and enrichment, with evaluation built in
• Define what good results look like for each use case: acceptable error rates, reference cases and what happens if a result is wrong
• Work with the Technical Product Manager and subject-matter leads to determine what is worth automating
• Documentation and tests are part of the work, not something done afterwards

Требования

• At least 4 years of experience as a Data Engineer or Data Scientist, with pipelines and automations that you built yourself and operated in production
• Confident with SQL and Postgres
• Confident using Python for data work: pandas, notebooks, evaluation scripts
• Experience integrating SaaS tools via APIs and webhooks, including two-way sync
• Practical experience with LLMs in data pipelines: embeddings, structured extraction, evaluation
• Experience with matching, classification or ranking methods and how to measure whether they work
• You check what an API actually does instead of relying on the documentation
• You consider GDPR from the outset: consent, retention periods, access
• You use Claude Code or similar AI coding tools daily and know where their output needs to be thoroughly checked
• Fluent in English
• Будет плюсом: Experience with record linkage or entity resolution; German is a plus

Условия

• A role with real responsibility: you will shape JUNI’s data platform and the data behind our tools
• Work that matters to science-based founders, researchers and our partners in Berlin-Brandenburg
• A motivated team with a hands-on mentality and flat hierarchies
• Flexible working hours and hybrid work
• A modern working environment in the heart of Berlin
• Compensation that reflects experience and responsibility