Blake Gilliland

Blake Gilliland

Team Lead, Data Scientist @ Home Partners of America

About Me

Contact: bagilliland5328@gmail.com

I am the leader of the Resident Success and Marketing work-stream at Home Partners of America. HPA is a portfolio company of Blackstone, based in Chicago, IL. We specialize in rent-to-own single-family homes as well as traditional rentals in several markets throughout the United States.

My team works closely with HPA’s senior leadership to develop analytically driven strategies, such as predictive machine learning models, prescriptive statistics, A/B tests, custom algorithms, and data visualization that will further the satisfaction and success of our residents. We care primarily about the resident experience and we look for opportunities within our lease operations to make life as good as possible for them. A sample of the high impact projects I’ve led:

  • Resident Partner Program - Conducted A/B test to measure impact on the resident experience of re- imagined ResComm strategy, resulting in CSAT 40%→ 60% & NPS 30 → 52.

  • Asana API Client - Designed, engineered, and productionalized Python API via AWS which connects legacy data with third party task management tool used by entire company, resulting in $1m+ recovered rental revenue, 1k+ man-hours gained, 10% reduction in resident touch-points.

  • Resident Experience Prediction - Classification models built to predict bad residents experience resulting in reduction of escalations by 10% and High Cost maintenance by $100k+.

  • Lease Renewal Prioritization - Regression Modeling employed to prioritize lease renewal execution op- portunities resulting in $500k+ saved rental revenue.

  • Telephony Routing Optimization - Custom algorithms developed to identify bottlenecks in call routing system resulting in Answer Rate (50% → 85+%) and Wait Times lowered (30 mins → 2 mins).

  • Resident Experience Segmentation - Implemented unsupervised clustering algorithms to segment resi- dents into ‘good’/‘bad’ resident experience clusters resulting in new company KR’s and resident success strategies, and tracked in an executive-level PowerBI report.

It is my responsibility to closely engage with my team, which includes several direct reports, so that they are empowered to boldly take on complex analytical problems and do so with a strong sense of ownership and collaboration. I do this by giving constructive, transparent, and candid feedback. We are open minded to new ideas and approaches that allow us to get as close to optimality as possible, without trading off time constraints. While some Data Science units may view themselves as ‘internal consultants’ to the business leaders at their company, we see ourselves as thought-partners who hold just as much stake in the success of our residents as the leaders of the company do. This makes collaboration much more meaningful for us and we thrive by having a close pulse on the operations in addition to the data coming out of them.

Before HPA, I served as the senior data scientist for Discover Financial’s DCX data product in their Payment Services division. This product generates some $400m to Discover each year. It is sold to external businesses that make use of merchant information, such as business names and addresses and MCC codes, to verify transactions from their own consumer base. My role was to ensure that the data was accurate and ready for production use. We did this by building statistical models to predict incorrect classes via ML algorithms such as boosted trees. Before Discover, I was apart of ALDI’s logistics department where I served as a data scientist in their supply chain team where my focus was on cost reduction via waste elimination strategies via predictive analytics including time series forecasts and regression analysis.

Skills

I use Python primarily for statistical modeling. See the below libraries I have experience with.

Data Mining: pandas, numpy

Visualization: seaborn, matplotlib

Cloud: boto3

Modeling: sklearn, tensorflow, keras, xgboost, lightgbm, pymc-marketing, statsmodels, prophet, scipy

Version Control: git

Command Line: bash

I have a lot of experience with R, and frankly I think it is an awesome language for scripting, but it can be slow and is not ideal for OOP or productionalization so I don’t as much as I’d like to. I’m an expert with SQL and prefer to use it as much as I can before doing analysis in Python or some other tool. I have a lot of experience with Tableau, which is my preferred dashboarding software. I also have some experience with PowerBI. Lambda, Fargate, and EC2 are the services from AWS I have the most experience with. I use Terraform for deployment.

On top of this, I have very strong leadership, communication, and relationship building skills. They have set me up for success both as a people leader in a data science organization as well as a business leader alongside executives looking to harness the power of data science to improve and understand their own processes.

Background

My background is in mathematics and applied statistics, particularly probability theory. I passed Exam P from the Society of Actuaries as an undergraduate. I then went on to graduate from FGCU with a 4.0 GPA in mathematics and statistics. My interests are primarily with Bayesian statistics, predictive modeling, and AB testing/Experimental design.