Role overview
- Automatic classification of real estate documents (floor plans, land registry records, identity documents).
- Structured data extraction from property files and contracts (Data Parsing).
- Automatic drafting of contextual client communications.
Basic qualifications
- Must Have
- LLMs in Production: Proven experience shipping AI features to production
- Agentic Workflows: Advanced skills in tool use, agent orchestration, and versioned prompt management.
- Evaluation Frameworks: Experience building eval pipelines on non-deterministic systems to measure hallucinations and accuracy.
- Privacy & Compliance: Demonstrable experience with data masking, PII scrubbing / anonymisation, and GDPR compliance when integrating commercial LLMs (e.g. zero-retention API policies).
- Nice to Have
- Interoperability experience between Python services and JVM backends (Java / Kotlin / Scala).
- Knowledge of RAG patterns, vector stores, and embedding pipelines.
- Familiarity with the Real Estate, Fintech, or vertical CRM/ERP domain
About the company
Casavo is an Italian proptech company operating in the residential real estate market. We build a vertical SaaS platform used daily by our real estate agents — a custom ERP and CRM developed in-house, serving real users and handling complex operational processes. After a radical optimisation phase, our tech team is now an elite nucleus of 6 engineers, with an extreme focus on pragmatism, code ownership, and a frameworkless architecture philosophy.