Data and Analytics Engineer

Nimble Gravity

About The Position

As Analytics Engineers, you would combine insurance domain expertise with full-stack data and analytics engineering capabilities. You will help build the data foundations that power Snowflake's AI platform. This role is focused on the layers that make AI reliable: clean, well-modeled data, governed pipelines, and semantic models that expose business meaning to natural language interfaces. You will design rigorous data models, build data pipelines, and construct the semantic layer that sits between raw data and AI agents. You will be responsible for setting up data that is structured, trusted, and agent-ready. The deployment patterns and data model gaps you surface feed directly back to LOB teams, making you both a practitioner and a source of signal for what gets built next.

Requirements

Advanced SQL: CTEs, window functions, incremental pipeline patterns. You can write complex queries without referencing documentation.
Analytics engineering and data modeling: Experience building data infrastructure involving large-scale relational datasets; strong instincts for pipeline design, QA, and testing across the full stack from ingestion through semantic layer.
dbt: Experience building and maintaining dbt projects with testing, documentation, and CI/CD pipelines.
Python: Modern, type-hinted, readable. You understand Python-based data pipelines and automation workflows.
AI-assisted development: You have used an LLM coding assistant (CoCo, Cursor, GitHub Copilot, Claude, or equivalent) as your primary development environment. Daily usage is the baseline.
Semantic modeling: You can write a semantic view configuration or structured skill file that handles edge cases and encodes enough domain knowledge that the model behaves like a subject matter expert.
Client-facing communication: You write code, but your output needs to make sense to a business leader who has never opened a terminal. You are the translation layer between what Snowflake's AI can do and what the customer actually needs.
Snowflake Cortex: Cortex Analyst, Cortex Agents, Cortex Search, semantic views, Dynamic Tables.
8+ years of experience in analytics engineering, data engineering, or a related technical role, with at least a portion of it customer-facing or cross-functional
Daily use of an AI coding assistant as a primary development tool
Proficient in SQL; can write window functions and complex joins without referencing documentation
Experience with dbt
Has shipped production data model or pipeline that non-technical business users actually relied on
Comfortable in Git (PRs, branches, code review)
Demonstrable experience translating business requirements into technical specifications

Nice To Haves

Experience with Airflow or other orchestration frameworks.
Familiarity with enterprise business systems (ERP, CRM, HRIS, or similar).

Responsibilities

Architect flexible, performant data models that drive LOB team toward single sources of truth across their LOB business domains
Use SQL, Python, dbt, and Snowflake to build and maintain data infrastructure for reporting, analysis, and automation
Perform data QA and develop automated testing procedures for Snowflake data models
Provide input into data governance strategies including permissions, data lineage, and data definitions
Design Data security so that the model only has access to the data
Build semantic data models that expose LOBs data to natural language queries via Cortex Analyst, turning complex schemas into something a business stakeholder and Clients can ask a question of
Define and validate the metrics, dimensions, and relationships that AI agents need to reason correctly over LOBs data
Identify and resolve gaps in data structure, naming, and coverage that would cause an agent to fail or produce incorrect results
Documented playbooks, reusable data model templates, and semantic model libraries that can be maintained and extend
Run technical workshops to upskill other team members
Author semantic view configurations and skill files (YAML + Markdown) that a non-technical analyst can invoke in plain English
Tracks adoption after go-live, identifies stall points, and re-engages until the data product is reliable and can be handed over to run teams.
Turn patterns into reusable templates and playbooks that the next engineer can deploy at the next customer, not to build bespoke every time.
Engages with customers to derive requirements, prototypes fast, gathers feedback, and iterates.
Distills messy deployments into clean, actionable feedback for Leaders, explaining root causes and suggesting fixes, not just reporting problems.