David Albrecht 🔮 Email & Phone Number

San Diego, CA, US

North Bethesda, MD, US

Working on recommendations, search relevance, and general ML.

Irvine, CA, US

• Implementing the first versions of the product’s recommendation systems, including product and brand content filtering and product hybrid filtering models.
• Improving user-facing search by defining and tracking success metrics for search, and iteratively improving the Elasticsearch system.
• Improved models for product category classification. Doubled the offline accuracy of the classifier from 40% to over 80%, and the average recall at 6 over 1,200+ categories to 86% using a multi-task learning.
• Maintain and optimize legacy ETL scripts. In one case, reduced the run time from 8 hours down to 30 minutes using parallelization.

San Diego, C, US

• Engaged in end-to-end data science projects from understanding business objectives to deploying solutions at scale. I construct datasets using Python and SQL to extract data from databases, APIs, and web.
• Leveraged ResNet50 and DistilBERT to create a user-facing product recommendation system based on more than 1.5 million products with user-provided images, free-form text, and structured categories.
• Enabled client to scale indefinitely by removing the need for human-labelling of topics to articles. Leveraged DistilBERT and Latent Semantic Indexing (LSI) to upgrade their current recommendation system.
• Combined ResNet50 and Locality Sensitive Hashing to create user-facing image similarity recommendation system from half a million images. Leveraged an Auto-Encoder (Deep Learning) to achieve an 80% compression in.

• Created scripts in Python to automate data retrieval and cleaning for analysis.
• Analyzed historical shipment claims against weather data to report on effectiveness of recommendations to clients. Presented results to the company.

• Worked on 2 projects: 1) Age Friendly Communities (commercial) - identifying communities in San Diego experiencing the greatest difficulty finding and affording assisted living care services; 2) Healthy Food Access.
• Collected and analyzed census data using Python, and conducted meetings to find appropriate data.

London, GB

• Worked on over 30 projects to assist audit team valuations and direct client valuations.
• Project subjects included: Contingent considerations, total shareholder returns, and convertible and straight debt.
• During busy season, worked on a new project pursuit regarding changing IRS guidelines to conversion ratio taxation. Worked between 70-100 hours a week.
• Created a project tracker in Excel that enabled me to efficiently keep track of 15+ projects at a time. Distributed to the national team.

London, GB

• Client facing role in both internal, business process audit and external, IT audit on 5 different projects. Performed SOX readiness testing, IT General Controls testing, and conducted client meetings.
• Experience with Journal Entry Data Analytics in ACL to support journal entry completeness testing by Assurance Services. Oversaw scheduling for projects and assist in training new staff.
• Planned multiple “counseling family” meetings, which enabled groups of coworkers to network and reconnect.
• Consistent role in the full-time in-house interview, peer/intern advisor, and Emerging Leaders Program processes.

Data Mining Certificate, Data Science

Data Science Career Track

Bachelor'S Degree, Finance And Financial Management Services

Education record

Education record