Total Experience Required: 3+ years
Mode of Work: Work from Office
Location: Hyderabad

About the Role :
We are looking for a Data Engineer to own and extend the Databricks Lakehouse that powers every Bay6 AI product – Scout, Rebound, Guide, Pathways and Verify. You will work at the center of our data architecture, ingesting client CRM/SIS data (Slate, Banner, EAB Navigate, Stellic, Salesforce, etc.) in the context of complex and potentially fragmented tech stacks inside of Higher Education institutions. You will typically build out a medallion pipeline (bronze/silver/gold) and maintain a multi-tenant Unity Catalog isolation framework across a growing client portfolio. You will partner closely with the engineering team and the AI solutions team to ensure that every agent has clean, governed and well modeled data to work on.

This is a hands-on role in a small, fast-moving setting where you will also be writing pipeline code and making architectural decisions. You must be comfortable working with clients and also be able to articulate decisions or concerns eloquently.





Bay6 AI operates as a headless solution with no native UI. All agent output write into client CRMs. Data accuracy and traceability aren’t trivial concerns as they show up directly in a client’s production CRM. You will be one of the few engineers who would deeply understand the full data path from a client’s SIS all the way to an agent decision, which means real ownership and visibility into product outcomes.

Must Have Skills
Databricks - Spark (PySpark or Spark SQL, Delta Lake (medallion architecture), Unity Catalog (Multi-tenant)
Python & SQL
Experience owning multi-tenant data architecture
Experience in Higher Education technology (CRM/SIS/Degree Audit platforms, etc.)
Working knowledge of FERPA, GLBA, SOC 2
Data Security and Governance
GitHub, working with a CI/CD
Architecture Reviews, decisions
Evaluating emerging tools, models and frameworks

Nice to Have Skills
Exposure to agentic AI system

📌 Data Engineer (Hyderabad)
🏢 BAY6.AI
📍 Hyderabad