Lead Data Platform and ML Manager
lighthouse life solutionsSpringfield (IL)
Lead Data Platform and ML Manager
Posted yesterday
lighthouse life solutionsSpringfield (IL)
SENIORITY
Manager
About the role
Lead Data Platform and ML Engineer
Company: LHL Strategies, Inc.
Location: Remote (United States), with periodic travel
Reports To: Chief Technology Officer
Department: Technology
About LHL Strategies:
Every year, millions of seniors lapse or surrender their life insurance policies for little or no value from their insurance company. Life settlements give them the option to sell the policy instead, for significantly more than they would otherwise receive. The money we pay helps meet retirement needs, healthcare costs, long term care expenses, or whatever else they choose. Lighthouse Life is part of LHL Strategies, Inc., a group of companies providing vertically integrated services for asset managers who acquire and manage life policies and portfolios. We are the only integrated platform in the industry spanning policy acquisition and trading, valuation, management, and portfolio servicing. We are a remote-first company, and this role can sit anywhere in the U.S. Our offices are in Conshohocken, Pennsylvania, just outside Philadelphia. We encourage innovation, protect work-life balance, and are driven to exceed expectations.
About the Role:
Our data runs on a Databricks lakehouse on AWS. It brings together structured and unstructured data from across the business and feeds analytics and machine learning models. This role owns that platform. The lakehouse and our first production models are already in place. You will own their ongoing operation and evolution: the pipelines that feed the models, the models themselves, and the evidence that keeps them defensible to regulators and auditors. The lakehouse holds data on behalf of multiple owners, and each owner sets its own terms for what we may collect, how it can be used, and who it can be shared with. Respecting those boundaries is a core requirement of the platform, not an afterthought. You will make sure that every pipeline, data product, and model reflects the agreed ownership, sharing, and ingestion rules, both in the work already built and in everything we add. You will also be able to show that this is true when someone asks. We are looking for someone who treats data stewardship as an engineering discipline and conforms to and builds on the controls we have in place. This is build and run, not build and hand off. The platform already matters, and you get to decide where it goes next: bringing in new data sources, extending the models we have, and building new ones. How We WorkWe are remote and distributed, and we operate with high transparency and an ownership mentality. Expect a daily standup, visibility into your work via Teams, and blockers flagged before they become emergencies. What we value:· Flagging problems early, not after they have become crises· “How can we improve this?” over “This is how we have always done it”· Systems thinking: automate and document, do not just execute manually· Intellectual honesty about what a model does and does not know If you thrive when told exactly what to do and left alone, this is not the role. If you thrive when trusted to own outcomes and keep leadership informed, we want to talk.
Roles and Responsibilities:
- Operate and improve the pipelines that bring data from our operational systems and document sources into the lakehouse· Extend the lakehouse to new internal and third-party data sources as the business grows, with quality checks and governance built in from the start· Enforce data ownership, sharing, and ingestion boundaries across every pipeline, data product, and model, and be able to demonstrate that they hold· Own the dbt transformation layers and the curated data products built on top of them· Govern the lakehouse in Unity Catalog: access controls, lineage, and the handling of sensitive data as new tables and consumers arrive· Own the platform as code, with infrastructure and workloads deployed through automated, reviewed CI/CD pipelines· Keep the data quality framework running: automated tests, freshness checks, and anomaly detection, held to an agreed reliability target· Own document ingestion and extraction using Databricks AI functions and Amazon Textract, and extend it to new document types over time· Extend our existing machine learning models and design, build, and ship new ones, from problem framing and feature design through validation, governance review, and production· Run the model lifecycle in MLflow: feature pipelines, retraining, evaluation, registry, versioning, and approval, kept auditable· Operate batch and real-time inference through Databricks Model Serving against the latency targets the business requires· Monitor models in production, respond to drift, and maintain the explainability, fairness testing, and documentation our model governance requires· Build and maintain the semantic layer and Databricks AI/BI dashboards our business teams use, reconciled to authoritative sources so one question does not produce two answers· Own cloud cost for the data and machine learning estate against an agreed budget, and keep runbooks and documentation current
Desired Skills:
- Hard Skills·
- Data engineering. 5 or more years building production data pipelines. Strong SQL and Python, production experience with Spark or PySpark, and dbt for transformation and testing.·
- Databricks. Hands-on production experience with Unity Catalog, Delta Lake, Lakeflow, Databricks SQL, and Databricks AI functions.·
- AWS and infrastructure as code. Running Databricks on AWS, with infrastructure as code in Terraform or OpenTofu and workloads as Databricks Asset Bundles.·
- CI/CD. Pipeline ownership using GitHub Actions, with automated testing and controlled promotion across environments.·
- Machine learning engineering. Building and shipping models end to end: feature pipelines, experiment tracking, MLflow, and both batch and real-time serving.·
- Model operations. Drift detection and response, retraining strategy, champion and challenger evaluation, and explainability methods such as SHAP.·
- Regulated data handling. HIPAA or equivalent. PHI and PII protection, encryption, least-privilege access, and audit logging.·
- AI-assisted development. We build with AI every day, and we expect our data and AI engineer to do the same. You use tools like Claude Code, Databricks Assistant, or GitHub Copilot as part of your normal workflow, and you review their output as carefully as a teammate’s.·
Helpful, not required. Document extraction at scale. Taking over and hardening a platform built by another team.
Soft Skills·
Owner. You treat the model and the platform as yours, not as something that belongs to whoever built them.·
Honest about uncertainty. You will say when a model is not good enough yet, including when that is inconvenient.·
Writer and communicator. We are remote and asynchronous, and much of this role is documentation that a regulator or auditor will read. Excellent written and verbal communication skills are required.·
Business-curious. You will want to understand how the business works, not only how the pipeline runs. You work closely with our Lead Developer and with business, analytics, and compliance teams.·
Pragmatic. A model that ships and is monitored beats a better model that is still in a notebook.
Education·
Bachelor’s degree in Computer Science, Statistics, Mathematics, Engineering, or a related quantitative field, or equivalent practical experience. A master’s degree is helpful. Insurance, actuarial, healthcare, or credit risk modeling experience is a plus.
Compensation and Benefits: ·
Base salary: Competitive, commensurate with experience·
Benefits: Comprehensive health, dental, and vision; 401(k); flexible PTO·
Remote work: Remote-first, anywhere in the U.S.Why Join LHL?·
Impact: Own the data and model platform for a growing company in an underserved market· Innovation: Work at the intersection of finance, insurance, and machine learning· Culture: Collaborative, entrepreneurial environment with experienced leadership
Before you apply
Applying takes about a minute. These four things decide how fast it moves after that.
Your profile is current
It's what we read first. Occupations, seniority and locations matter more than a long history.
Two examples you can talk through
Not a portfolio — just two pieces of work where you can explain the decisions and what you'd change.
A number in mind
What you're on now and what would make you move. We negotiate better when we know both.
Your notice period
Employers plan around it, and it's the question that stalls offers most often.
Once you apply, someone reads it and calls you before anything reaches the employer — usually within two working days.