We are looking for a curious, solution-oriented AI prompt engineer who loves leveraging modern AI tools to solve real-world mapping and data problems. You must know how to „vibe code” - using tools like Cursor, Claude etc. to turn ideas into working Python scripts, test the outputs, and build end-to-end automation pipelines.

You will combine basic geographical concepts with AI workflows to help streamline key data operations across our ecosystem, supporting the development of exploratory Jupyter notebooks and data analysis pipelines.

The Day-to-Day Activities

• Prompt Design & Architecture: Craft, test, and iterate on system prompts, few-shot examples, and chain-of-thought workflows to extract geospatial attributes (e.g., road types, speed limits, barriers, etc.).
• Automated Map Quality Control: Build LLM-powered evaluation and test-data generation pipelines to handle automated map-quality validation and offload manual QA workflows.
• Geospatial Benchmarking & Evaluation: Establish rigorous evaluation frameworks using ground truth datasets, precision, recall, and spatial error taxonomies to measure and improve prompt accuracy over time.
• Pipeline Integration & AI-assisted development: Use Python, LLM APIs (Claude, ChatGPT), and development agents (Cursor) to integrate structured prompt outputs into automated, scheduled execution pipelines (e.g., Prefect) and build interactive Jupyter notebooks.
• Collaborate cross-functionally with data scientists, ML engineers, and map experts to integrate LLM outputs with geospatial analysis and gather requirements.
• Analyze patterns and anomalies in map data using statistical and ML techniques.
• Share findings via visualizations and presentations that influence product or research directions, and drive documentation and reproducibility standards.

Requirements:

• Prompt Engineering Expertise: You will have 0–2 years of experience, including internships, academic projects, or early-career work, designing system prompts, enforcing structured outputs (JSON/schemas), handling edge cases, and optimizing LLM performance for production workflows.
• Model Evaluation Literacy: Strong foundation in evaluation metrics, including ground truth curation, precision, recall, false-positive/negative analysis, and LLM output benchmarking.
• Technical & Scripting Skills: Working knowledge of Python, including pandas, NumPy, matplotlib/seaborn, and scikit-learn. Hands-on experience using Python for prompt testing, working with APIs, and developing scripts or Jupyter Notebooks through iterative, AI-assisted development.
• Machine Learning Foundations: Solid understanding of basic ML concepts such as classification, regression, and clustering.
• Experience working with LLM orchestration frameworks (LangChain, LlamaIndex, etc.).
• Analytical Mindset: Excellent problem-solving skills with a drive to replace manual map operations with creative, scalable AI solutions.
• Effective communication skills (written and verbal), with the ability to present insights to technical and non-technical audiences.
• Strong collaboration skills, a proactive attitude, and a willingness to continuously learn, adapt, and give/receive feedback constructively.

Good to Have:

• Familiarity with Computer Vision or geospatial data analysis.
• Knowledge of basic DevOps or cloud platforms (e.g., Azure, AWS, GCP).
• Prior experience working in mapping, mobility, or logistics domains.