JR and SR AI Developer

Protective Life is transforming how it builds and operates software — moving to a product operating model organized around empowered, outcome-oriented teams — and is putting machine learning and generative AI to work in the products that serve our customers. Voyager is one of these product pods, spanning our Life, Annuities, and Employee Benefits lines.

As a JR or SR AI Developer on Voyager, you help build the AI-powered features and services that reach real users — integrating large language models and ML into Voyager's products on our Databricks Lakehouse and Microsoft Azure. This is a hands-on individual-contributor role focused on application engineering with AI. You will build well-scoped features with guidance from senior engineers and the AI/ML Engineering Lead, working closely with product managers, designers, and data engineers. As a regulated life insurer, we expect AI features to be accurate, well-documented, and appropriate in their handling of sensitive customer data.

 

• Build AI-powered application features and services on Azure Databricks and Azure — integrating LLMs and ML models into Voyager's products, with guidance on design from senior engineers.
• Implement GenAI capabilities — retrieval-augmented generation (RAG), embeddings and vector search, prompt and system design, and tool/function calling.
• Develop and consume APIs and services that expose model capabilities to product surfaces, with attention to latency, reliability, and cost.
• Apply evaluation, guardrails, and human-in-the-loop review to keep AI outputs accurate, safe, and appropriate for a regulated insurer.
• Work with the pod's data stack — dlt (dltHub), dbt, and Dagster — to source and prepare grounding data and features for AI capabilities.
• Deploy and version the models and prompts your features use with MLflow and Databricks Model Serving, following patterns set by the AI/ML Engineering Lead.
• Write clean, tested, version-controlled code and ship it through Azure DevOps (ADO) CI/CD.
• Instrument AI features for monitoring — output quality, latency, cost, and user feedback — and help iterate based on evidence.
• Apply secure-by-default and privacy practices for sensitive customer and policyholder data used in AI features — PII handling, access control, and data minimization in prompts and context.
• Collaborate with product managers and designers to refine AI features through discovery and iteration.
• Contribute to responsible-AI and governance practices — evaluation evidence, documentation, and adherence to model/AI governance expectations.
• Grow your craft — seek and apply feedback in code and design reviews, and share what you learn with the pod.
 
SENIOR:
• Design and build AI-powered application features and services on Azure Databricks and Azure — integrating LLMs and ML models into Voyager's products.
• Develop GenAI capabilities — retrieval-augmented generation (RAG), embeddings and vector search, prompt and system design, tool/function calling, and agentic workflows.
• Build and consume APIs and services that expose model capabilities to product surfaces, with attention to latency, reliability, and cost.
• Implement evaluation harnesses, guardrails, and human-in-the-loop review to keep AI outputs accurate, safe, and appropriate for a regulated insurer.
• Integrate with the pod's data stack — dlt (dltHub), dbt, and Dagster — to source and prepare grounding data and features for AI capabilities.
• Deploy and version the models and prompts your features depend on using MLflow and Databricks Model Serving, in partnership with the AI/ML Engineering Lead.
• Write clean, tested, version-controlled code and ship it through Azure DevOps (ADO) CI/CD.
• Instrument AI features for monitoring — output quality, latency, drift, cost, and user feedback — and iterate based on evidence.
• Apply secure-by-default and privacy practices for sensitive customer and policyholder data used in AI features — PII handling, access control, and data minimization in prompts and context.
• Partner with product managers and designers to shape AI features through discovery and rapid, evidence-based iteration.
• Contribute to responsible-AI and governance practices — documentation, evaluation evidence, and adherence to model/AI governance expectations.
• Mentor less-experienced engineers and share applied-AI patterns and reusable components across the pod.

• 3–5 years of software development experience, including hands-on work building AI-powered or GenAI applications.
• Solid programming skills — Python required; familiarity with JavaScript/TypeScript or a JVM language a plus — with sound software-engineering fundamentals (APIs, services, testing).
• Practical experience building GenAI/LLM features — RAG, embeddings and vector search, prompt/system design, and basic evaluation.
• Experience integrating models via APIs and model-serving platforms — exposure to Azure OpenAI and Databricks Model Serving / MLflow preferred.
• Familiarity with the modern data stack the pod uses — dlt (dltHub), dbt, and Dagster — on a Databricks lakehouse (Delta Lake); willingness to grow here.
• Experience with CI/CD (Azure DevOps / ADO preferred) and Git-based, test-supported development practices.
• Working knowledge of a cloud environment (Microsoft Azure preferred), including AI/OpenAI services basics.
• SQL and comfort working with data.
• Attention to evaluation, documentation, and secure, compliant handling of sensitive data.
• Bachelor's degree in Computer Science, Engineering, or a related field — or equivalent practical experience.

PREFERRED QUALIFICATIONS

• Experience in financial services or insurance products (Life, Annuities, claims, servicing, or customer experience).
• Experience with agent and orchestration frameworks (e.g., LangChain, LlamaIndex, or Semantic Kernel) and vector stores (Databricks Vector Search or Azure AI Search).
• Front-end or full-stack experience delivering AI features into user-facing products.
• Familiarity with responsible-AI and evaluation tooling, and with bias/fairness and explainability considerations.
• Relevant certification such as Microsoft Azure AI Engineer Associate or a Databricks GenAI/ML credential.
 
SENIOR

REQUIRED QUALIFICATIONS

• 5–8 years of software development experience, including recent, hands-on work building AI-powered or GenAI applications.
• Strong programming skills — Python required; familiarity with JavaScript/TypeScript or a JVM language a plus — with solid software-engineering fundamentals (APIs, services, testing).
• Hands-on experience building GenAI/LLM applications — RAG, embeddings and vector databases, prompt/system design, tool/function calling, and structured evaluation.
• Experience integrating models via APIs and model-serving platforms — Azure OpenAI and Databricks Model Serving / MLflow preferred.
• Experience working with the modern data stack the pod uses — dlt (dltHub), dbt, and Dagster — on a Databricks lakehouse (Delta Lake).
• CI/CD experience with Azure DevOps (ADO) and Git-based, test-supported development practices.
• Working knowledge of Microsoft Azure — compute, storage, identity, and Azure AI/OpenAI services.
• Strong SQL and comfort working directly with data.
• Demonstrated attention to evaluation, documentation, and secure, compliant handling of sensitive data.
• Bachelor's degree in Computer Science, Engineering, or a related field — or equivalent practical experience.

PREFERRED QUALIFICATIONS

• Experience in financial services or insurance products (Life, Annuities, claims, servicing, or customer experience).
• Experience with agent and orchestration frameworks (e.g., LangChain, LlamaIndex, or Semantic Kernel) and vector stores (Databricks Vector Search or Azure AI Search).
• Full-stack or front-end experience delivering AI features into user-facing products.
• Familiarity with responsible-AI and evaluation tooling, and with bias/fairness and explainability considerations.
• Familiarity with model risk and governance expectations in regulated settings.
• Relevant certification such as Microsoft Azure AI Engineer Associate or a Databricks GenAI/ML credential.