Senior AI Engineer (all genders)

Das ist der Job

The role is fully remote, focusing on Germany and Serbia.

Darum lohnt es sich

What You’ll Do Discover and Define AI Opportunities Work directly with Product, the VP of Engineering, the Engineering Team Lead, and compliance experts to understand pain points and manual workflows that AI can automate. Document your systems clearly so the team can understand, operate, and extend them.

Stay current with the rapidly evolving AI landscape (model providers, tooling, evaluation frameworks) and bring relevant developments to the team. Benefits High ownership, direct impact on growth, and clear development opportunities; annual training budget; trusted ideas.

Additional perks: Urban Sports Club membership, Germany Ticket, company pension plan, and other benefits. Our team thrives on the unique perspectives and experiences that each person brings.

The more diverse and inclusive we are as a team, the greater our work will be. #J-18808-Ljbffr heyData is building a compliance platform that guides companies — from 10‑person startups to 400‑person multi‑entity organisations — step by step to audit‑ready outcomes across ISO 27001, NIS2, GDPR, and more.

Compliance is fundamentally a knowledge and process problem, and AI is one of the most powerful tools we have to solve it. Role Overview We’re hiring our first Senior AI Engineer to identify where AI can eliminate manual work, build the systems that make it happen, and own those systems end‑to‑end.

You’ll work directly with Product, Engineering, and our compliance experts to discover real problems and ship real solutions. Translate vague problem descriptions into clear AI feature specs: what the system does, what data it needs, how success is measured. Push back confidently when AI is not the right solution — and propose alternatives.

Build and Ship AI‑Powered Features Design and implement AI features using foundation models (OpenAI, Anthropic, Mistral) as a starting point, customised with RAG pipelines, prompt engineering, and fine‑tuning where justified. Expose your AI services as clean RESTful APIs consumed by the rest of the application (NestJS/VueJS stack on AWS EKS).

Build data ingestion and transformation pipelines to feed AI systems with heyData's internal data and customer data, handled with strict GDPR compliance. Integrate AI agents and multi‑step workflows where appropriate, using frameworks like LangChain or LlamaIndex pragmatically.

Own the AI Lifecycle Manage the full lifecycle of every AI system you build: development, deployment, monitoring, and retraining. Set up CI/CD pipelines for AI models integrated into our existing AWS infrastructure. Monitor model quality in production — detect drift, evaluate output quality, and trigger retraining when needed.

Experiment and Measure Design and run AI/ML experiments rigorously: clear hypotheses, appropriate evaluation metrics, statistically sound interpretation. Communicate findings to non‑technical stakeholders in plain language.

Qualifications 5+ years total software engineering experience, with at least 3 years specifically building and operating production AI/ML systems. Proven track record shipping LLM‑powered features to real users.

Hands‑on experience with major foundation model APIs: OpenAI, Anthropic Claude, Mistral; familiarity with platform abstractions like AWS Bedrock or Google Vertex AI. Production experience with RAG architectures: vector databases (Pinecone, Weaviate, pgvector), embedding model selection, chunking strategies, retrieval evaluation.

Solid Python engineering: production‑grade APIs and services; comfortable building services that integrate cleanly into a microservices architecture. Experience with CI/CD for AI systems: model versioning, experiment tracking (MLflow, Weights & Biases or similar), monitoring, and automated retraining pipelines.

Skills Strong problem decomposition: design an AI system that solves a vague pain point. Statistical thinking: design experiments properly, interpret results rigorously, and know when findings are meaningful.

Clear communication: explain model trade‑offs, architecture decisions, and experiment results to engineers, product managers, and non‑technical stakeholders. Data pipeline fluency: build data ingestion and transformation infrastructure, and understand useful vs noisy data.

Values Trustworthy: ship what you commit to, and communicate early when something changes. Ownership‑driven: own your systems from prototype to production monitoring. Curious: actively explore problem spaces with stakeholders, asking questions until you understand the real problem.

Pragmatic: choose the simplest solution that works; avoid unnecessary infrastructure. Collaborative: work well with engineers and domain experts. Remote, in Berlin office, or hybrid — choose where you’re most productive. 30 vacation days, flexible working hours, and a work culture that respects focus time.

Top equipment: MacBook & more to help you get started smoothly. Equal Opportunity At heyData, we’re proud to be a vibrant and diverse startup. We celebrate and embrace people from all backgrounds, including diverse gender, age, sexuality, religion, ethnicity, disability status, parental status, and more.

We’re committed to fostering an inclusive and supportive environment where everyone can succeed.