Senior Machine Learning Engineer (Modeling), Risk

Company Description

Since we opened our doors in 2009, the world of commerce has evolved immensely, and so has Square. After enabling anyone to take payments and never miss a sale, we saw sellers stymied by disparate, outmoded products and tools that wouldn’t work together.

To solve this problem, we expanded into software and built integrated solutions to help sellers sell online, manage inventory, book appointments, engage loyal buyers, and hire and pay staff. Across it all, we’ve embedded financial services tools at the point of sale, so merchants can access a business loan and manage their cash flow in one place. Afterpay furthers our goal to provide omnichannel tools that unlock meaningful value and growth, enabling sellers to capture the next generation shopper, increase order sizes, and compete at a larger scale.

Today, we are a partner to sellers of all sizes – large, enterprise-scale businesses with complex operations, sellers just starting, as well as merchants who began selling with Square and have grown larger over time. As our sellers grow, so do our solutions. There is a massive opportunity in front of us. We’re building a significant, meaningful, and lasting business, and we are helping sellers worldwide do the same.

Job Description

As a Machine Learning Engineer within the Risk Machine Learning and Decision Science team, you work on projects that enable a software driven, machine learning centric view on all money movement and every transaction within the rapidly growing Square ecosystem. This touches on actively maximizing the trade off of revenue growth and risk using artificial intelligence. The machine learning driven software that we release interacts with every transaction and money movement within our seller ecosystem - a profound degree of scale and impact.  Such machine learning techniques touch on reinforcement learning, decision theory, deep learning sequence modeling, natural language processing, and optimization theory.  In addition, we also strive to provide our sellers, through seller facing products, with transparency around why our machine learning made a particular decision.  This touches on algorithms in the relatively new space of explainable artificial intelligence.  

Our algorithms derive value from our unique and rich data from our entire product portfolio within our rapidly growing seller ecosystem. We partner with business, product, operations, and engineering teams to drive optimal decision making systems using sophisticated modeling and machine learning. We’re a passionate team of entrepreneurs, scientists, and engineers who are shipping machine learning software that actively actively manages Square’s view on each transaction as it pertains to our revenue growth and risk.  

You will: 

• Build machine learning/deep learning models that analyze payment activity in real time across our Seller’s ecosystem consisting of payments, banking, and debit card products. 
• Adapt existing machine learning methods and transfer learning to develop solutions that work at global scale.
• Leverage an experimentation mindset along with state-of-the-art algorithms to create preventative systems, collaborate on new product features to drive losses down, and explore new datasets (including 3rd party data) to engineer new features for our models. 
• Collaborate with business leaders, subject matter experts, and decision makers to develop success criteria and optimize new products, features, policies, and models
• Research, design, develop, and test a range of classification, regression and optimization problems

Qualifications

You have:

• An advanced degree (M.S., PhD.), preferably in Computer Science,Engineering, Statistics, Physics, Mathematics or a related technical field.
• PhD plus 3 years (or Master plus 5 years) industry working experience in applied Machine learning or Deep learning 
• A strong track record of performing machine learning model development using Python (numpy, pandas, tensorflow, pytorch, scikit-learn, etc.) and SQL/NoSQL interaction patterns.
• Expert level knowledge of modern techniques in machine learning and deep learning, e.g., tree models, transformer network architectures, with an orientation to maximizing such algorithms in a large scale production setting. 
• Familiarity with Linux/OS X command line, version control software (git), and general software development principles with a machine learning software development life-cycle orientation.
• Machine learning strategic sequencing of methodological and software improvements to work back from maximizing core metrics associated with optimizing the business.
• The ability to clearly communicate complex results to technical and non-technical audiences and stakeholders (PMs, Operations, Engineers).

Additional Information

Block takes a market-based