Job Description
PropertyExpert is the technology and development partner of the insurance and real estate industry. The concept for success is based on the intelligent linking of man and machine: digital processes and AI-based document verification combined with human expertise.
Based on millions of data records from the document review, our technology experts develop modular end-to-end and predictive solutions around the claims process, enabling the efficient and at the same time customer-oriented handling of property damages. Our mission is clearly defined: to work together with clients to get smarter - claim by claim.
- Train deep learning models for information retrieval, classification und generation tasks
- Visualize and exlpore insurance claims data, such as repair invoices, formulars, images etc.
- Apply GenAI techniques for document understanding tasks
- Deploy trained modes to production using Google Cloud Platform tooling
- Keep improving models/prompts through better data and code
- Out stack: Python, PyTorch, Transformers, GPTs, SQL, mongodb, FastAPI, Docker, GCP, dash, streamlit
- Masters Degree in Computer Science or other quantitative discipline
- Solid understanding of machine learning algorithms & principles
- At least three years of experience with Python (Scikit-learn, Pandas, Numpy)
- At least two years of experience designing, implementing, and training deep NN architecture
Desired skillset
- Degree with a focus on deep learning and NLP tasks would be a plus
- Ability to write code based on published papers would be desirable
- Experience in training transformers models from scratch (incl. pre-training)
- Experience in self-supervised machine learning
- Familiarity with creating standalone microservices
- Experience with public cloud infrastructure (GCP, AWS or Azure)
- Experience with data visualization using dash or streamlit
The current Research & Development Team consists of 6 team members
- A long-term position with a certified family-friendly, financially sound employer – something that has not changed in the pandemic
- Competitive pay, occupational retirement benefits, supplementary health and accident insurance that is fully employer-financed, bike-leasing
- A structured onboarding process and a personal mentor
- 30 days of vacation, flexible working times, self-managed working time, and working time accounts to ensure an optimum work-life balance
- Options for working from home/mobile working
- A real team spirit, flat hierarchies, an open-door policy, and an open error management culture