AI Engineer III

Base Salary Range USD $133,836.99 - USD $150,000.00 /Yr. Join Mariner Finance!

Since 1927, the Mariner Finance family of companies has provided customers with creative, flexible, and convenient lending options. Headquartered in Baltimore, Mariner Finance operates coast-to-coast with physical locations in over half the states. With a growing number of employees, superior customer service remains the cornerstone of our business, and we pride ourselves in delivering a variety of loans with an enhanced focus on exceptional service. We work with customers to find options that are beneficial to their specific needs, which is why we are recognized by our customers as one of the community’s consumer finance companies of choice.

Benefits:

For information regarding our benefits, please visit: https://www.marinerfinance.com/careers/benefits/

All full time employees are provided with a generous benefits package in addition to their monetary compensation. Learn more about it today.

Job Details

In this role, you will...

Be responsible for building and maintaining the knowledge layer and backend services that power enterprise AI agents. Develop the skills and tool registries, knowledge graph components, vector search indexes, data integration pipelines, and agent evaluation suites that help turn standalone models into reliable, governed, tool-using agents.

Partner with Level 4 engineers, the AVP of AI Engineering, Data, Security, and business stakeholders to deliver well-scoped components of the platform. Apply established engineering patterns, contribute to design discussions, and grow technical depth in agentic AI and knowledge engineering.

Responsibilities and Duties:

• Build and maintain components of the enterprise knowledge layer that agents use for grounding, memory, and context, including skills, knowledge graph elements, and vector indexes, working from designs set by senior engineers and contributing improvements.
• Develop and maintain parts of the skills and tool registry, including skill and tool definitions, versioning, discovery, and publishing workflows, and build MCP servers and APIs that expose capabilities to agent builders.
• Implement knowledge graph pipelines, including entity extraction, entity resolution, relationship loading, and schema updates, and build graph and hybrid retrieval queries for agent use cases.
• Build and tune vector search, including embedding generation, semantic chunking, metadata filtering, hybrid search, and index refresh jobs, and measure retrieval quality against agreed metrics.
• Build batch and streaming data pipelines on Databricks (Delta, Spark, Structured Streaming) and modern ETL/ELT patterns to ingest structured, semi-structured, and unstructured data into the knowledge layer, with data quality checks and lineage.
• Build integrations between business systems, document stores, and the knowledge layer using secure REST APIs and connectors. Register and govern data and AI assets through Unity Catalog.
• Implement access controls and entitlement checks in retrieval paths so agents return only data that a given user and agent are permitted to see.
• Build agent evaluation suites, including golden datasets, synthetic test cases, LLM-as-judge and rule-based scorers, and regression tests that measure retrieval quality, groundedness, tool-selection reliability, task completion, safety, latency, and cost.
• Integrate evaluations into CI/CD pipelines as quality gates and review evaluation results to identify regressions and improvement opportunities.
• Write clean, asynchronous, production-ready Python and robust web APIs (such as Flask or FastAPI) that power agent actions and expose knowledge services. Build and integrate agentic tools and workflows for Databricks Custom Agents.
• Apply context and prompt engineering techniques to manage context windows and produce structured outputs (such as JSON schemas) from tool-calling agents.
• Implement monitoring, tracing, and logging for agent loops, tool usage, and retrieval systems, and follow security practices for identity, secrets, runtime isolation, and data store protection.
• Contribute to CI/CD pipelines, test triggers, and multi-environment deployments, deploy services on AWS (including EC2), and use infrastructure as code where available. Participate in code reviews and follow Git branching and merge practices.
• Create and maintain technical documentation, runbooks, and lineage records that support auditability, supportability, and responsible AI governance.
• Troubleshoot and resolve issues in pipelines, retrieval quality, integrations, and agent behavior, escalating complex issues to senior engineers as needed.
• Partner with engineers, data owners, and business stakeholders to understand requirements and deliver reliable components. Learn from code reviews and technical guidance, and share knowledge with peers.
• Perform additional duties as assigned to support evolving business needs.

Required Qualifications:

• Bachelor’s degree in Computer Science, Engineering, Mathematics, Statistics, Information Systems, or a related field; additional, applicable years of experience may be substituted for a bachelor’s degree.
• Minimum of six (6) years of experience in software engineering, data engineering, platform engineering, machine learning engineering, or a related technology discipline, with hands-on experience in AI, machine learning, or LLM-based solutions (production work, substantial projects, or formal training).
• Strong Python programming skills, including asynchronous programming and testing, and experience building REST APIs (such as Flask or FastAPI).
• Experience building data pipelines using ETL/ELT patterns and SQL, preferably on Databricks or a comparable platform such as Snowflake.
• Working experience with vector databases or vector search, embeddings, chunking strategies, and retrieval-augmented generation (RAG) pipelines.
• Working knowledge of knowledge graphs or graph databases (such as Neo4j, Amazon Neptune, or RDF/SPARQL) and graph or ontology modeling concepts.
• Experience or demonstrated familiarity with building evaluation datasets and automated scoring for LLM or agent outputs.
• Familiarity with agentic tooling, including tool calling, structured outputs, and agent frameworks, and awareness of Model Context Protocol (MCP) or similar tool-integration approaches.
• Familiarity with monitoring, tracing, and logging for applications and distributed services.
• Experience integrating services with business systems through secure APIs, with working knowledge of identity, secrets management, and access control.
• Experience with AWS (including EC2), CI/CD pipelines, and Git branching and code review practices.
• Working knowledge of responsible AI, data privacy, security, and auditability practices.
• Ability to work independently on well-defined tasks, ask for guidance when needed, and collaborate with engineers, data owners, and business partners.
• Strong written and verbal communication skills, with the ability to explain technical work to technical and non-technical audiences.
• Strong analytical thinking and problem-solving ability, with a willingness to learn new tools and techniques quickly.

Preferred Qualifications:

• Master’s degree in Computer Science, Artificial Intelligence, Data Science, Mathematics, Statistics, Engineering, or a related technical discipline.
• Exposure to the Databricks ecosystem, including Mosaic AI Agent Framework, Vector Search, MLflow tracing and evaluation, Databricks SQL, and Unity Catalog.
• Hands-on exposure to agentic frameworks (such as LangGraph or similar), multi-agent coordination, and automated tool calling.
• Exposure to GraphRAG, entity resolution, and graph query languages (such as Cypher, SPARQL, or Gremlin).
• Familiarity with hybrid search, re-ranking, and embedding model evaluation.
• Experience building MCP servers or reusable tool and skill definitions for agents.
• Understanding of model selection, token and inference cost optimization, and embedding models.
• Experience with infrastructure as code (such as Terraform) and containerization (such as Docker).
• Experience with AI or data solutions in a regulated environment such as financial services, banking, lending, or insurance.

Hours of Work:

Work hours will depend on the business hours of the time zone serviced. To the extent permitted by law, the Company may, in its sole discretion, change the work schedule to address business needs.

Physical Demands:

While performing the duties of this job, the employee is frequently required to sit for extended periods; reach with hands and arms; and talk or hear. The employee is occasionally required to move about. The employee must occasionally lift and/or move up to twenty (20) pounds. Specific vision abilities required by this job include close vision and the ability to adjust focus.

EEO:

Mariner Finance is an Equal Opportunity Employer and does not discriminate on the basis of race, color, religion, creed, sex, gender, gender identity or expression, marital status, age, religion, national origin, sexual orientation, familial or caregiver status, citizenship status, status as a victim of domestic violence, medical condition, genetic information, pregnancy, physical or mental disability, or status as a disabled or Vietnam era veteran. Employee must be able to perform the essential duties/functions of the position satisfactorily and, if requested, reasonable accommodations will be made to enable employees with disabilities to perform the essential duties/functions of their job, absent undue hardship. Drug/Alcohol/Smoke-free workplace.

This job description is not designed to cover or contain a comprehensive listing of activities, duties, or responsibilities that are required of the employee. Duties, responsibilities, and activities may change or new ones may be assigned at any time or without notice.