Job Purpose
To design and deliver AI-powered Business Intelligence solutions by integrating Databricks Genie, Semantic Models, and Metric Views on top of a robust end-to-end data engineering stack.
The role focuses on enabling self-service analytics using natural language (GenAI), ensuring business-friendly data consumption, and building governed, scalable, and high-performance data ecosystems across batch and real-time pipelines.Duties and Responsibilities
KEY ROLES / PRINCIPAL ACCOUNTABILITIES
AI for BI & GenAI Enablement
• Design and implement GenAI-powered BI solutions using Databricks Genie or MS Fabric
• Create and manage Data Rooms for business users with curated datasets
• Define and maintain Instructions Layer (prompt engineering for business context)
• Enable natural language to SQL/insights workflows for self-service analytics
• Drive adoption of Databricks One & other platforms as a unified analytics interface
Semantic Layer & Metrics Engineering
• Design and manage Semantic Models for business abstraction
• Develop reusable and governed Metric Views (KPIs, aggregations,



business definitions)
• Ensure consistency across BI tools (Power BI / Genie)
• Align semantic layer with business glossary and data governance policies
Data Engineering & Platform Development
• Build scalable pipelines using Azure Databricks (PySpark, SQL)
• Develop and orchestrate ETL workflows using Azure Data Factory
• Work with Delta Lake architecture (Bronze–Silver–Gold layers)
• Enable real-time and batch data processing pipelines
• Enable data exposure via APIs for BI and downstream systems
CI/CD & DevOps
• Implement CI/CD pipelines for data and AI workflows
• Automate deployments across environments (Dev, QA, Prod)
• Ensure version control and reproducibility

?
KEY RESPONSIBILITIES
• Translate business problems into AI-driven BI solutions
• Own end-to-end delivery: ingestion ? transformation ? semantic layer ? AI consumption
• Design scalable data architecture aligned with lake

📌 Data Engineer (Pune)
🏢 Bajaj Finance
📍 Pune