Product Data Scientist

Job Description
*About Pyyne*
Pyyne is a modern technology consultancy engineering the next generation of digital products and services.

At Pyyne, we believe in using technology to unlock business potential, create sustainable growth, and drive forward digital excellence. Our solutions range from advanced Software Engineering, Cloud, and Data & AI solutions.

Job Summary
Our client is a prominent NYC-based health and wellness company that provides weight loss coaching and medications. They are seeking a mid-level
*Product Data Scientist*
to conduct product analysis and help design & evaluate A/B tests.

The Product Data Science team currently consists of 4 other Data Scientists, who work together to tackle a variety of projects each week. They sit alongside the Business Data Science team (also 4 people) and the BI Engineering team (which is responsible for the data pipelines and dashboards they use).

"Product" at the client covers all their in-app programs and features, as well as their growth efforts (i.e. how we position and price their different product offerings). The stakeholders they work with most often are the various Product Managers (PMs) in the Product org.

Key Responsibilities
*Product Analysis:*

• Answer insightful analysis questions from stakeholders across the Product org
• Be able to ramp up on how different parts of the product & their data model function, to pull accurate results
• Partner with PMs and Engineers to design key metrics for different initiatives

*Experiment Analysis:*

• Be the "stats expert" alongside PMs to ensure we design effective experiments
• Understand and be able to intuitively explain concepts like power & sample size
• Analyze experiment results to determine what is statistically significant, being able to take into account common pitfalls and "p-hacking" that can happen with A/B testing

*Specific Deliverables (First 3-6 months):*

• Ramp up on a particular product area and start answering ad hoc insight questions from there
• Design an experiment (with another Product DS there to mentor), see it launch, and then analyze it
• Give feedback on new team processes and come up with an idea to iterate / make a new one

Your Day-to-Day might look something like this:

*Analysis (~60% of the job)*

• Be able to explore, query, and transform complex data using SQL and Python
• Answer open-ended insight questions from the data
• Apply basic descriptive stats, summarize findings, make visualizations of the data, etc

*Experimentation (30%)*

• Design "experiment plans" for new product & feature launches
• Solidly understand the statistics behind sample size and power, to design a valid A/B test
• Analyze experiments by running statistical tests on the results (i.e. t-tests) and interpret the output (p-values, including applying corrections for multiple comparisons)

*Machine Learning (10%, for now)*

• Be able to apply basic machine learning techniques (i.e. regressions and decision trees) to aid in analysis

• The business uses ML quite a bit elsewhere - financial forecasting, operational forecasting/predictions/etc

• Out of scope at the moment is doing ML development / evaluation for AI in the product

• But it is part of the PDS team scope and may be something this individual does, or grows into with mentorship

• The client is embracing an AI-first culture - this has just started so while ML for PDS may not be a major component currently, we expect this to grow quite a bit over time

Qualifications

• 3 - 5 years of experience as a data scientist, in a product-focused environment
• Strong proficiency in SQL and a solid understanding of data modeling concepts
• Proficiency with Python, ideally data analysis/science packages like Pandas and Stats
• Good "Data storytelling" and stakeholder communication: can explain technical concepts in an intuitive way
• Detail-oriented: can sense-check your output and correct errors before sharing results with stakeholders
• A good product-sense and familiarity with fundamental "product" metrics (like DAU, retention, etc)
• General understanding of machine learning, and how to make and evaluation models