AI Engineer (Need 14+ years exp)
Job Title: AI Engineer Location: Remote Hire-Type: Long-term Contract Technologies: SQL, Azure Stack, Python LLM prompt engineering Implementing OCR as a service is an Azure-native, real time event-driven scalable document OCR pipeline that processes multi-page documents in parallel using Azure stack solutions and docTR. Building an AI agent that processes insurance claim PDFs through multi-modal extraction (text, layout, images), generates comprehensive summaries, integrates with line-of-business and transactional databases to surface discrepancies, detect claim bias, and evaluate demand packages — with interactive querying for adjusters throughout. Sql, Azure Stack, Python LLM prompt engineering The Data Engineer will support the Data Science team by building and maintaining Azure-based data and document-processing pipelines. This role spans ingestion through extraction and model-ready transformation, enabling advanced analytics, classification, and decision-intelligence workflows. Core Responsibilities:
- Azure Pipeline Development:
– Build scalable ingestion, preprocessing, classification, and extraction pipelines using Azure Data Factory, Azure Databricks, ADLS, and Azure Functions.
- Document Processing:
– Implement OCR, NLP, and LLM-based extraction workflows using Cognitive Services or custom Databricks models.
- Orchestration & Runtime Management:
– Manage workflow orchestration, monitoring, and error handling with ADF, Databricks Workflows, and Azure Monitor.
- Data Science Enablement:
– Deliver clean, structured, model-ready datasets using Delta Lake and Databricks. – Support Azure ML deployments and integrations with downstream Q&A and retrieval systems.
- Integration & Decision Intelligence:
– Integrate processed outputs with decision-intelligence layers, search systems, and analytics dashboards (e.g., Power BI, Azure Search).
- Scalability, Governance & Best Practices:
– Ensure pipelines meet Azure security, governance, and performance standards. – Implement CI/CD, versioning, and automation for reliable, production-grade workflows.