Alexander Hasha Email and Phone Number

Advancing Mission Lane's machine learning automation and model risk management strategy by implementing frameworks and platforms for automated model development, validation, and monitoring.- Delivered a credit line assignment model to drive 15-20% increase in present value per booked account, with an estimated profit impact of $20-30 million per year.- I architected the project as a fully automated and containerized training and evaluation pipeline that can be regularly refreshed via Airflow DAGs running in Kubernetes.- Developed project templates and training to drive adoption of this development architecture across the Data Science team.

Persefoni is a leading carbon accounting SaaS firm. I led the engineering pod responsible for building the data and software architecture to implement calculations of greenhouse gas emissions aligned to the GHG Protocol for corporate emissions and the PCAF standard for financed emissions.- Developed ETL pipelines and database structures to integrate third-party emission factors with the calculation engine- Introduced documentation and testing processes that prevented calculation errors and reduced key-person dependencies- Developed algorithms to compute proprietary emission factors and to recommend emissions calculation methods based on the users available data

I defined requirements and prioritized the backlog for three engineering pods responsible for Capital One's first successful enterprise cloud platforms for model development, analytics, and batch model deployment, enabling thousands of internal analysts to harness the power of open source in the public cloud.

I led the model validation and risk management for machine learning models in diverse business domains, including chatbots, cybersecurity, data management, marketing optimization, hiring, identity verification, and fraud. My team upheld standards of practice, ensured regulatory compliance, and enabled informed executive decision making through independent evaluation of machine learning technologies.

I led a modeling team with the mission to prove out the potential of open source and machine learning:- Increased accuracy of Bundle’s ML matching algorithms to 99% precision, meeting accep- tance criteria for roll-out to all Capital One customers’ digital transaction ledger.- Operationalized the model as Capital One’s first Hadoop application and first production Python application.- Developed a software framework for building and deploying ML pipelines that was adopted by multiple teams.- Evangelized open source collaboration methods and software engineering best practices across Capital One’s Data Science community.

I led a team of developers and data scientists building consumer-facing data products based on billions of credit card transactions.- Developed a machine learning pipeline to power Bundle’s core product, a merchant review website- Hired and led a team to scale the data product and bring it to market- Worked with CEO to pitch company for acquisition, and led the technical due diligenceprocess in the company’s acquisition by Capital One

Mortgage Analysis, Markets Quantitative Analysis

Mathematics PhD Student