- Design and build a centralized semantic data layer using Cube Core (or equivalent technology) that provides a unified, governed abstraction over all company data sources.
- Define semantic models, metrics, dimensions, and relationships that map to business domains.
- Expose the semantic layer via REST/GraphQL APIs and MCP-compatible tool interfaces for consumption by AI agents and LLMs.
- Integrate and unify data from heterogeneous systems including MySQL, DynamoDB, Aerospike, Snowflake, Amazon S3, Apache Kafka, Amazon SQS.
- Build connectors, adapters, and federation layers to query across operational (OLTP) and analytical (OLAP) data sources.
- Design tool interfaces and API contracts that allow AI agents to discover data, understand schema semantics, and generate accurate queries autonomously.
- Collaborate with AI/ML teams to optimize the semantic layer for LLM-generated SQL, natural language querying, RAG, and agentic workflows.
- Architect the semantic layer with native multi-tenant isolation, ensuring strict data segregation and tenant-scoped access controls.
- Implement row-level security, column-level masking, and attribute-based access controls (ABAC) to enforce data governance policies.
- Design for horizontal scalability to support thousands of concurrent queries from AI agents, internal dashboards, and customer-facing products.
- Serve as the technical authority on data architecture decisions, authoring ADRs and reference architectures.
- Mentor and guide senior engineers on best practices for semantic modeling, data governance, and API design.
GraphQLPostgreSQLSQL+8 more