The Company
Perch is transforming the way people buy and sell their homes. Simplifying it all, to the way it should have always been; fair and true to market, straightforward, easy.
Every year in the U.S., $1.8 trillion of single family residences transact, generating over $80 billion of brokers fees in a process that has changed little in decades. For the average American, the home purchase and sale process takes months, creates anxiety and is filled with uncertainty and hassle. Perch offers a modern alternative, making one of life's biggest decisions -- the sale and purchase of a home -- less intimidating.
Perch is headquartered in New York City and has 100+ employees. We have raised over $50 million in financing from top tier investors including: FirstMark, Accomplice and Juxtapose.
The Team
Perch’s Data Science team is responsible for building the predictive models that help our underwriters make intelligent decisions when valuing homes. We use our skills in statistics, applied mathematics, computer science, and software engineering to value off-market real estate as accurately as possible. We also develop algorithms that power Perch's consumer search capabilities, using computer vision to identify home features and enable a new kind of home discovery.
Our goal is to incorporate first-party and third-party data in order to continually refine and improve our home valuations over time, using techniques as simple as linear regression and as complex as neural networks, whatever delivers the best result. This role is very performance-oriented: the accuracy of our models very directly impacts the profitability and success of our business.
Role and Responsibilities
We are currently seeking a driven, results-oriented Data Science Manager for our team. The ideal candidate is pragmatic and has great instincts for which modeling changes are most likely to bear fruit. Most Importantly, the Data Science Manager must be able to drive the team to obtain the best possible performance and is unabashed in providing feedback and even engaging in conflict if need be in order to make the team successful. The existing team is small (two data scientists), so it is expected that the Manager would be a player-coach, spending about 75% of their time making substantive individual contributions. The candidate should be accustomed to delivering results in hours and days rather than weeks or months.

This position reports to the Head of Investments & Data Science.

Professional Qualifications
5+ years of data science experience required
Prior team management experience required
Entrepreneurial - comfortable talking to stakeholders to understand business needs, running small tests to validate assumptions, and refining requirements based on results
Self-motivated and gets stuff done
Understanding of probability theory and statistical inference - required
Proficiency in Python - required
Knowledge of SQL - required

Experience with neural nets/computer vision - preferred
 
Personal Qualifications
Results orientation with a high motor and an incredible attention to detail; able to drive projects from planning to completion with limited oversight
Superior analytical skills and attention to detail with a data-driven mentality
A low ego and can-do attitude; willingness to admit mistakes and work to remedy them
Demonstrated communication and interpersonal skills to work with diverse stakeholders and cross-functional teams
Flexibility to prioritize deliverables and re-prioritize them at a moment’s notice
Entrepreneurial mindset--you see ambiguity as opportunity and obstacles as chances to build new solutions