Staff Applied AI Engineer

Build end-to-end, production AI systems that connect unified structured and unstructured data to decision-grade analytics and agentic LLM solutions, with architectural ownership across the enterprise.

Responsibilities

• Unify structured and unstructured data by creating pipelines that bring together Snowflake datasets (plus adjacent warehouses and lakes) and unstructured sources including text, documents, logs, and transcripts into modeling-ready datasets.
• Deliver decision-grade analytics, including customer churn analytics with quantified uncertainty (confidence or credible intervals) and clear communication on what the metrics support and where assumptions limit interpretation.
• Build predictive and prescriptive models that move from forecasting to action by developing propensity and optimization or recommendation systems that drive concrete business decisions.
• Engineer agentic AI systems by designing and shipping LLM workflows and agents that are token-efficient through context management, retrieval and caching strategies, model routing, and evaluation harnesses that protect cost and latency without sacrificing quality.
• Architect for the enterprise by defining reference architectures, integration patterns, and governance standards across ingestion, model development, MLOps/LLMOps, security, and observability, supported with diagrams and documentation.
• Own quality and reliability by establishing evaluation, monitoring, and guardrails for drift, accuracy, bias, safety, and cost across both classical ML and GenAI systems.
• Partner across the business by translating ambiguous business needs into technical solutions and explaining technical tradeoffs to non-technical stakeholders.

Requirements

• Hands-on data engineering experience with Snowflake (including modeling, performance, and cost management) plus SQL, along with experience wrangling unstructured data.
• Applied statistics skills to build churn or retention models and express uncertainty correctly using confidence or credible intervals, including a clear understanding of underlying assumptions.
• Demonstrated experience building predictive and prescriptive analytics that shipped and influenced decisions.
• Production experience with LLM and agentic systems using frameworks such as LangGraph, Claude Agent SDK, CrewAI, or custom orchestrators, with a track record of optimizing token efficiency, cost, and latency.
• Production RAG experience is strongly expected at senior+ level, including chunking, hybrid search, reranking, and retrieval evaluation.
• Architecture strengths, including the ability to design and document end-to-end systems and patterns others can build on, with evidence supporting design choices.
• Strong Python and a software engineering mindset covering testing, version control, and CI/CD.
• Excellent written and verbal communication with comfort working asynchronously in a distributed team.

Technologies

• Snowflake
• SQL
• Python
• LangGraph
• Claude Agent SDK
• CrewAI
• RAG
• CI/CD
• AWS Solutions Architect
• Google Cloud Professional ML Engineer
• Azure AI Engineer
• TOGAF
• vLLM
• TensorRT

Benefits

• Friendly flexible working model supporting work-life balance, whether working from home or in the office.
• Competitive compensation and total rewards, including health, wellness, and financial plans for you and your family.
• Global, diverse teams with collaboration across 23+ countries.
• Learning and development with access to best-in-class learning tools and programs.
• Equity and belonging, with an emphasis on valuing every voice.

Nice to Have (Preferred)

• Cloud certifications (AWS Solutions Architect, Google Cloud Professional ML Engineer, Azure AI Engineer) and/or TOGAF for enterprise architecture.
• Experience with inference optimization such as quantization, model routing, caching, and vLLM or TensorRT-style serving.
• MLOps/LLMOps tooling experience and platform-building.
• Domain experience in your industry, and prior experience owning AI strategy or build-vs-buy decisions.

What Success Looks Like (First 6–12 Months)

• Deliver a unified data foundation combining Snowflake and unstructured sources for downstream modeling.
• Launch a churn analytics offering trusted by the business, including quantified uncertainty and clear recommended actions.
• Ship at least one production agentic solution that reduces token spend and latency versus a naive baseline while meeting quality requirements.
• Publish and get adoption of a documented reference architecture and standards across teams.

About the Role

• Full-stack data and AI builder role spanning raw data ingestion, defensible analytics, and production AI systems.
• Architect and implement both the analytics and agentic systems, while defining the reference architectures, patterns, and standards the organization can build on.
• Emphasis on architectural ownership and systems that are cost-efficient by design.