Hexa
Lead Data Engineer @Panora (backed by Hexa)
2w ago
EuropeLeadllmdata pipelinesdata ingestiondata normalizationdata enrichmentdocument processingobservabilityevaluation pipelines
Lead Data Engineer role to build and scale AI-powered systems for Panora's insurance broker AI products.
About the product
- Built for an AI-first, production-grade product:
- Languages: Python for AI, data and automation; TypeScript for backend services and product integrations
- AI systems: LLM-powered agents, structured extraction, evaluation frameworks, tracing, observability and feedback loops
- Infrastructure: AWS, with a serverless and managed-services approach designed for reliability and scale
- Data: MongoDB, document-processing pipelines, structured extraction workflows, internal datasets and evaluation datasets
- Integrations: Microsoft 365, CRM and ERP tools, insurer extranets and other systems used daily by insurance brokers
- Engineering environment: a large, evolving production codebase: where improving, refactoring and scaling existing systems is as important as building new ones
- What matters most is your ability to design robust systems, work with real-world constraints, and learn fast.
Nice to have
- Experience with evaluation systems, feedback loops or AI observability
- Experience in fintech, insurance or other regulated environments
How to apply
- 30min phone screen with Presci (Talent team)
- 30min interview with Fabian (Hiring Manager)
- 1h30 technical interview with Fabian through a peer-programming session
- 30-min interview with Diane (Co-founder & CEO)
- Onsite meeting with a member of the Product / Design team and Mat (Partner at Hexa)
- Reference checks & offer 🎉
- Hexa is committed to creating a diverse environment. All qualified applicants will receive consideration for employment irrespective of gender, origin, identity, background and sexual orientation.
- We know there’s a long way to go regarding diversity in our industry, which is why we encourage all applicants- especially those listed above- to apply to our open positions.
Other
- Panora is building a suite of AI-powered agents for insurance brokers - a highly regulated and operationally complex industry.
- Our ambition is to become the AI Operating System for brokers in Europe : automating high-friction workflows while improving compliance, advisory quality, and client relationships.
- The product is already live with brokerage firms across France and Belgium, with AI assistants in production (quotation automation, contract comparison, coverage analysis, compliance checks, document generation…).
- Panora enables brokers to save hours every week, reduce analysis time by up to 70%, and focus on what truly matters: advising clients .
- Founded by Diane du Paty (ex-VC, operator) and Fabian Langlet (repeat founder, AI product engineer), Panora is backed by Hexa (Aircall, Spendesk, Front).
- We are entering a key phase of scaling and industrializing our AI platform .
- At Panora, AI is not a feature - it is the core product layer , deployed in real-world conditions with strong constraints on reliability, traceability, and compliance .
- We’re looking for a Founding Data Engineer to help build and scale the systems powering Panora’s AI products.
- Your work will sit at the intersection of:
- AI systems: LLMs, agentic workflows, evaluation pipelines, tracing and observability
- Data engineering: ingestion, normalization, enrichment, extraction and document-processing pipelines
- Internal datasets: building high-quality, structured insurance datasets from policies, quotes, underwriting questionnaires and business rules
- Product & backend engineering: designing robust APIs, data models and scalable services used directly in production
- Insurance expertise: translating real-world insurance workflows, contract language and decision rules into usable data systems
- As the third engineering hire, you'll work directly with Fabian (CTO & Co-founder) and Jeremy (Founding Software Engineer), with strong ownership and direct impact on both product and technical direction.
- Join an existing, large-scale codebase and quickly develop a deep understanding of the systems powering Panora.
- Improve, refactor and scale critical parts of the platform while contributing new capabilities where they create the most impact.
- You'll work with complex workflows, unstructured data and real production constraints.
- You'll own problems end-to-end, from AI systems and data pipelines to customer impact.
- Success is measured by product impact, reliability and customer outcomes.
- Design, ship and improve AI-powered workflows used daily by insurance brokers
- Build evaluation, feedback and monitoring systems to continuously improve performance
- Turn complex insurance workflows into reliable AI-powered products
- Build and maintain data pipelines powering our products
- Process and structure unstructured data (contracts, emails, insurer documents)
- Improve the quality, reliability and observability of our systems
- Own systems end-to-end: from design and implementation to deployment and monitoring
- Contribute to architecture and key technical decisions
- Help define how AI, data and engineering scale at Panora
- We’re looking for a builder who ships, enjoys solving difficult problems, and thrives in a small, highly collaborative team.
- Experience building, operating and improving production systems in a startup or product-driven environment
- Experience working closely within an engineering team and contributing to a shared codebase
- Strong Python and backend engineering skills
- Experience with data-intensive products, AI systems, LLM applications or applied machine learning
- Comfortable working with messy, incomplete and unstructured real-world data
- Able to quickly understand, improve and scale existing systems—not just build greenfield projects
- High standards for reliability, data quality, maintainability and customer impact in production
- Comfortable taking ownership, moving quickly and making progress in ambiguous environments
- Strong product mindset: you care about solving meaningful customer problems, not just shipping technical features