Sr Data Scientist, People Analytics

Lennar•Waterford, FL

5d•Onsite

About The Position

We are Lennar Lennar is one of the nation's leading homebuilders, dedicated to making an impact and creating an extraordinary experience for their Homeowners, Communities, and Associates by building quality homes and providing exceptional customer service, giving back to the communities in which we work and live in, and fostering a culture of opportunity and growth for our Associates throughout their career. Lennar has been recognized as a Fortune 500® company and consistently ranked among the top homebuilders in the United States. A Career that Empowers You to Build Your Future As a Senior Data Scientist on Lennar's People Analytics team, you'll build the models that anticipate what's ahead for our workforce: attrition and turnover risk, hiring and headcount needs, and the trends underneath them. You'll also build the AI agents that deliver that work directly to HR leaders, so forecasts reach the point of decision rather than sitting in a report. This is a newly created position and the first dedicated data science role on the People Analytics team. You'll have significant latitude in shaping how the work gets done, from the modeling approaches we adopt to the standards we hold for validation and how results are presented to senior leadership. You'll work closely with HR COEs and hands-on with Workday and other people systems, plus the metrics Visier surfaces on top of them. You'll report to the Head of People Analytics.

Requirements

Bachelor’s degree or higher in Statistics, Economics, Math, Computer Science, or a related analytical field with equivalent experience.
5+ years of relevant experience (1+ with PhD, 3+ with MS) as a data scientist producing models and engineering solutions in a production environment.
Depth in predictive modeling, including regression, time series forecasting, and classification.
A strong statistical foundation, including experimentation and causal inference, and the judgment to select the right method for the question at hand.
Proven ability to work independently, including designing your own validation approach and articulating your methodology to technical and non-technical reviewers.
Hands-on experience building with LLMs, whether agents, tool use and function calling, retrieval, or prompt-driven workflows, and using tools such as Claude to accelerate your own development. We're interested in what you've built and how you evaluated it, more than years of experience in a field this new.
Experience with large, complex operational or HR datasets, including cleaning and structuring messy data and extracting it yourself through SQL or APIs.
Fluency in SQL and strong proficiency in Python.
Ability to explain model results and agent behavior in plain language to non-technical HR and business leaders.

Responsibilities

Partner directly with HR leaders to understand what they need forecasted or automated and build models and agents they trust enough to act on.
Build and deploy predictive models in Python and SQL.
Monitor model performance over time, retesting as new data arrives and refining when accuracy drifts, so predictions and recommendations stay reliable.
Build AI agents that can query People data, reason over it, and take action, along with the tool use, orchestration, and guardrails that let them work safely with sensitive workforce data.
Select the method that fits the question, apply experimentation and causal inference where the question is about cause, quantify uncertainty, and be prepared to explain the approach to senior leadership.
Use LLMs to surface patterns in unstructured People data such as survey comments and exit interview notes and incorporate what you find into your models.
Develop deep familiarity with Lennar's people data, extracting and structuring what your models require, identifying gaps that would sharpen the output, and partnering with Data Engineering on the pipelines behind it.
Explain statistical concepts, model results, and agent behavior to HR and business audiences without a quantitative background.
Handle sensitive workforce and compensation data with discretion.