Conversational BI Engineer

The Conversational BI Engineer designs, builds, evaluates, and operates governed natural-language analytics and AI
experiences for the company. The role connects certified enterprise data and semantic context to secure conversational interfaces
across Microsoft, Snowflake, Databricks, and SAP ecosystems.
Mission
Deliver accurate, explainable, secure, and supportable conversational BI experiences that preserve metric consistency, source
transparency, privacy, and human accountability.
Key outcomes (what success looks like)
 Production-ready conversational BI grounded in certified data products and governed semantics.
 Measured answer quality through repeatable evaluation of SQL accuracy, grounding, authorization, relevance, robustness,
latency, and cost.
 Transparent citations, assumptions, limitations, clarification behavior, and escalation paths.
 A repeatable lifecycle for intake, pilot, validation, deployment, adoption, monitoring, and support.
Responsibilities
 Configure trusted datasets, semantic instructions, business rules, sample and verified queries, synonyms, clarification
flows, citations, and response guardrails.
 Build experiences using Snowflake Cortex Analyst/Search/Agents and AI Functions, Databricks Genie/AI-BI and Unity
Catalog semantics, Microsoft Copilot Studio/Fabric/Azure AI, and SAP BDC/Joule or analytical AI capabilities.
 Implement RAG, semantic search, embeddings, vector stores, tool/function calling, structured outputs, agent orchestration,
identity-aware retrieval, prompt/version management, and human escalation.
 Create golden question sets, labeled test data, automated regression evaluations, adversarial/security testing, telemetry,
and quality dashboards.
 Apply RBAC, DLP, privacy, retention, logging, auditability, model and data access controls, safety policies, and Responsible
AI governance.
 Monitor adoption, failures, hallucination/grounding risks, semantic drift, latency, capacity, token/compute consumption, and
platform changes.
 Partner with subject-matter experts, semantic engineers, BI developers, cybersecurity, governance, legal/privacy, and
platform teams for production acceptance.
Required qualifications
 4–7+ years in BI, analytics engineering, AI solution engineering, application development, data science, or a related field.
 Hands-on delivery using at least one of Snowflake Cortex AI, Databricks Genie/AI-BI, Microsoft Copilot Studio/Fabric/Azure
AI, or SAP BDC/Joule, plus working knowledge of the other ecosystems.
 Strong SQL and proficiency with Python, APIs, JSON, authentication, semantic models, and enterprise integration.
 Experience with RAG, evaluation frameworks, prompt/instruction engineering, observability, security testing, CI/CD, and
production support.
 Ability to translate business questions into governed analytic intents, testable expected answers, and safe response
behavior.
Preferred qualifications
 Manufacturing, SAP, Finance, Supply Chain, HSE, or operations analytics experience.
 Agent frameworks, MLflow, model serving, vector search, Streamlit, notebooks, and embedded analytics.
 Responsible AI risk assessment, red teaming, privacy engineering, and human-in-the-loop design.
Expanded technology requirements
 Cross-platform architecture across Microsoft Fabric and Power BI, Snowflake, Databricks, and SAP Business Data Cloud,
including workload placement, interoperability, semantic consistency, identity, networking, data movement, resiliency, and
total-cost tradeoffs.
 SAP Business Data Cloud ecosystem: SAP Datasphere, SAP Analytics Cloud, SAP Databricks, SAP HANA Cloud, SAP
Master Data Governance, SAP BW modernization, curated business data products, and governed third-party connectivity.
 Snowflake ecosystem: architecture and virtual warehouses, RBAC, secure sharing, Snowpark, Streamlit, Cortex AI
Functions, Cortex Analyst, Cortex Search, Cortex Agents, embeddings/vector patterns, observability, and consumption
controls.
 Databricks ecosystem: lakehouse and medallion architectures, Delta Lake, Unity Catalog, Lakeflow, SQL Warehouses,
notebooks, MLflow, Model Serving, vector search, AI/BI dashboards, Genie Agents, and governed business semantics.
 Microsoft ecosystem: Power BI and Fabric semantic models, Direct Lake, OneLake, lakehouse/warehouse, pipelines,
notebooks, Copilot Studio, Azure AI services, APIs, deployment pipelines, monitoring, and capacity management.
 Engineering and governance: advanced SQL, Python, REST APIs, OAuth/service principals, secrets management, Git and
CI/CD, automated testing, metadata, lineage, data quality, FinOps, privacy, cybersecurity, and Responsible AI controls.
 Platform-neutral design using approved open formats, APIs, reusable data contracts, portable business definitions, and
documented integration boundaries to limit avoidable lock-in.
 Conversational platform depth: Snowflake Cortex Analyst, Search, Agents and AI Functions; Databricks Genie Agents,
AI/BI dashboards and Unity Catalog semantics; Microsoft Copilot Studio, Fabric and Azure AI; SAP BDC, SAP Analytics
Cloud and Joule-related analytical experiences.
 AI lifecycle: use-case qualification, prompt and semantic design, offline/online evaluation, release gates, monitoring,
incident response, cost management, model or platform change assessment, and continuous improvement.
Core competencies
 User-centered conversational design.
 Systematic accuracy, safety, and evaluation mindset.
 Evidence-driven experimentation.
 Transparent communication of limitations, risk, and value.
 Product ownership across adoption, reliability, cost, and supportability.