Samuel Kim Email & Phone Number

San Francisco, CA, US

Leading AI-driven data classification initiatives

Menlo Park, California, US

Experimentation & Metrics Platform (Tech Lead)- Revamped analysis platform to deliver analysis scorecard results in near-real time- Architected self-serve metrics repository to simplify metrics onboarding and centralize datasets- Scaled experiment data pipelines to handle hundreds of running experiments bymigrating from Presto to Spark- Scaled assignment.

Stanford, CA, US

Responsible for leading recitation sections, grading and writing homework, and holding office hours for CS courses on algorithms, natural language processing, probability, and discrete mathematics.

• Built data pipelines and trained classifiers to predict how likely admins of Facebook groups would accept or reject auto-reported spam posts
• Reduced number of false positives by 120K posts per day
• Increased recall by escalating 30K reports to deletions and feedlimits per day

• Fully decomposed the current Aspera Files provisioning process and reimplemented it as an automated workflow in ServiceNow
• Cut down manual overhead from up to 6 hours to several minutesImproved existing, developed and designed new microservices to provide APIs to support provisioning workflowEnhanced Aspera customer usage portal, improving.

Berkeley, CA, US

• Researched current methods for string graph assembly of genomes
• Theorized a ‘not-so-greedy’ approach and helped proved its feasibility under sufficient conditions
• Wrote a full implementation of the algorithm in C++
• Produced graphs comparing n50 scores that show huge improvements over current methods
• Contributing author to published paper

Mountain View, CA, US

• Implemented a full pipeline that tracks and publishes metrics for the throttling system in the Ads Budgeting team
• Wrote scripts to extract and calculate metric prediction stats daily that were run on jobs managed by Google’s machine cluster management system
• Published metrics to a dashboard and experimented with updated weights associated with these predictions using machine learning tasks

Mountain View, CA, US

• Designed and created an activatable UI metrics tool with a simple pluggable interface for Ads applications using AngularDart
• tool records and reports client-side page information
• tool records Angular specific information; i.e: cache sizes, slowest digests
• tool is used by two Google Ads applications that use AngularDart

Berkeley, CA, US

• Lead discussion and lab sections for 30 students in our lower-division Signals and Systems course
• Wrote up lecture notes, discussion worksheets, and homework solutions for the class in Latex
• Managed and coordinated all of the readers who grade homeworks and provided feedback to improve the course throughout the semester

Berkeley, CA, US

Spent time grading student's homework for the Intro to Signals and Systems class. In addition, I wrote practice problems using Latex to prepare students for their midterms.

Berkeley, CA, US

Grading student's homeworks and projects for CS61A, the introductory cs course taught in Python

Berkeley, CA, US

I assisted students in the lab for the introductory course taught in Python

San Francisco, CA, US

• Hired as the first intern for their startup
• Implemented much of the APIs, both server-side and client-side, for the company’s dashboard in AngularJS
• Implemented and design of users’ account settings and management
• Created necessary database constraints using postgreSQL and wrote the majority of their j-unit tests

Master Of Science (M.S.), Computer Science

Education record

Diploma