The Data/Ingestion Engineer will design and implement scalable ingestion pipelines for high-volume structured and unstructured insurance documents, including PDFs, scans, emails, Word, Excel and PowerPoint files.
The role includes OCR/document extraction, text cleaning and normalization, semantic chunking, metadata tagging, vector storage/RAG preparation, validation, monitoring and integration with downstream AI solutions.
Requirements
- At least 7 years of experience as Data Engineer with Python and SQL
- At least 5 with AWS, the rest should be more or less but should be real experience that the candidate is able to talk about and give a concrete example/
- Must to have :
- Python
- SQL
- AWS
- S3
- Step Functions
- CloudWatch
- Data/document ingestion pipelines
- Processing of unstructured documents:
- scanned documents
- Word
- Excel
- PowerPoint
- emails
- OCR / document extraction technologies
- AWS Textract or equivalent
- Connectors/integrations with sources such as:
- SharePoint
- Text cleaning and normalization
- Metadata extraction/tagging
- Automated error monitoring
- Validation of OCR/extraction results
- Git
- CI/CD
- Testing
- Public-cloud data processing/document ingestion experience
- Good to have :
- Vector databases
- RAG
- Semantic chunking
- Vector-storage schema design
- Retrieval mechanisms
- Azure
- Databricks
- AWS Textract specifically
- Experience within insurance / financial services
- Experience handling enterprise security and low-latency SLA requirements
- English Level C1
- Duration of the mission : 1 year