[Remote] Principal AI Platform Engineer
Note: The job is a remote job and is open to candidates in USA. Nextdata is a company focused on making enterprise data usable for both humans and AI agents. The Principal AI Platform Engineer will design and build interfaces and systems that allow AI systems to effectively discover and use data products while ensuring safety and compliance with governance policies.
Responsibilities
- Design agentic data interfaces that let AI agents discover, understand, and safely use data products
- Build MCP-compatible endpoints, tools, and APIs for governed AI/data access
- Develop data agents that reason over metadata, semantics, contracts, policies, and data outputs
- Make data products AI-ready across SQL, documents, vectors, graphs, APIs, and semantic models
- Build safe query and action flows with access checks, policy enforcement, approvals, and audit trails
- Work on retrieval, semantic search, tool selection, context construction, and answer grounding
- Define reusable patterns for agent-readable metadata, structured outputs, observability, and evaluation
- Partner with product, engineering, and customer teams to turn enterprise AI/data use cases into product capabilities
Skills
- Strong experience in data engineering, data platforms, distributed systems, or enterprise data infrastructure
- Practical experience building AI-enabled data systems, retrieval systems, semantic layers, or data agents
- Strong knowledge of SQL, APIs, documents, vector search, knowledge graphs, and metadata systems
- Experience with agentic interfaces, tool-calling, MCP or similar protocols, function calling, or AI backends
- Good understanding of governance: access control, policies, contracts, lineage, data quality, PII protection, and auditability
- Ability to build production systems that are safe, observable, testable, and reliable
- Strong Python skills and comfort working across backend services, data systems, APIs, and AI frameworks
- Product-minded judgment: you know the difference between a demo, a customer-specific workaround, and a reusable platform capability
- Comfort working in ambiguous areas where the patterns are still being defined
- Experience with data mesh, data products, semantic models, catalogs, governance platforms, or data marketplaces
- Experience with MCP servers, tool registries, LLM orchestration, RAG systems, or multi-step agents
- Experience with Databricks, Snowflake, BigQuery, Spark, DuckDB, Postgres, graph databases, vector databases, or lakehouse architectures
- Experience with enterprise identity and authorization systems such as SSO, OAuth, OIDC, SAML, SCIM, RBAC, ABAC, or policy engines
- Experience evaluating AI systems for retrieval quality, tool-use accuracy, groundedness, reproducibility, and failure modes
Company Overview