Sr. AI/Machine Learning Engineer

Sr. AI/Machine Learning Engineer
Department IT and Programming

Employment Type Full-Time, Remote (Memphis, TN candidates preferred for in-office collaboration)
Minimum Experience Experienced
Role Summary

This is a builder’s role, not a research role. You will write the Python that puts AI models to work on real production problems: reading messy documents and email, resolving entities across systems, enriching records, scoring likelihood, and surfacing signals that were previously invisible.

We are looking for an engineer with good working knowledge of transformer architecture and practical experience with foundation models on both sides of the market: open-source models you can host and run, and commercial models you consume through an API. You do not need to have trained one from scratch. You do need to be comfortable calling them, prompting them well, handling their output, and building reliable services around them.

The role also spans traditional machine learning. We have a large and interesting data set, and part of the job is finding the modeling opportunities hiding inside it that translate into better recovery outcomes: classification, matching, scoring, and prediction. You will also help with synthetic data approaches where real data is limited or contractually restricted. This is a remote position; candidates in or near Memphis, TN are preferred.
Core Responsibilities

- Write clean, production-quality Python that integrates foundation models into automated pipelines and services, similar to our existing document intake, routing, entity resolution, and data enrichment workflows.

- Work with both open-source and commercial foundation models, including prompt design, tool calling, structured output, error and retry handling, and evaluating which model fits a given workload on accuracy, latency, and cost.

- Uncover and shape modeling opportunities in our data that lead to stronger recovery outcomes, then build them: classification, entity matching, ranking, and propensity or likelihood scoring.

- Design, train, evaluate, and deploy machine learning models using standard modeling and automated machine learning platforms.

- Build retrieval and multi-step model workflows using orchestration frameworks, including state handling and guardrails.

- Help develop synthetic data approaches where real data is sparse, sensitive, or contractually restricted, including generation strategy and validating that the synthetic data actually improves model performance.

- Support the machine learning operations layer: training and inference pipelines, model versioning, deployment automation, and monitoring for drift and performance.

- Integrate models into production applications and workflows through APIs and services, so models land in the product rather than in a notebook.

- Build practical evaluation into everything you ship: test sets, before-and-after comparisons, human review where it matters, and honest reporting of failure modes.

- Optimize models and services for performance, scalability, and cost, including inference and token consumption.

- Spot opportunities in the data while organizing chaos and cutting through noise, and speak up when the right answer is something simpler than a model.

- Follow responsible AI and data handling practice: PHI protection, access controls, model documentation, and traceability of what a model was trained on.

Qualifications
Experience

- 5+ years in machine learning, data science, or data engineering, including experience putting models into production use.

- Strong proficiency in Python and SQL. You should be comfortable writing and maintaining the integration code yourself.

- Good working knowledge of transformer architecture and how modern foundation models behave.

- Practical experience with both open-source and commercial foundation models, such as Llama, Mistral, or Qwen alongside Anthropic Claude or OpenAI, including prompt design, tool use, and structured output.