Brian Okoro

chibuezebrokoro@gmail.com
Long Beach, CA 90804
909-247-6376

Industry, Research, and Volunteer Experience

Bird Rides

Sr. Operations Analyst → Data Analyst → Associate Data Scientist
  • Worked primarily under Director of Data Science Dave Herman, with time under Principal Data Scientist Tittu Thomas
  • Staged gradient-boosted tree model to identify the best new locations to place vehicles to ride, building an interactive map-based Tableau product on top. Viewed by downstream stakeholders over 400 times in 6 months, it projected a 67% lift in utilization compared to new locations selected manually by Operations and general management
  • Developed an exponential smoothing-based model for hourly opening of locations to release vehicles. This model achieved the same error as the principal scientist’s incumbent model, but ran weekly instead of hourly, saving 99.4% in compute resources
  • Conducted exploratory analysis identifying opportunity 46% lift in time vehicles spent between getting damaged by conducting preemptive repairs during vehicle’s first repair session, leading to preemptive repair as a primary focus of the company’s 2020 vehicle maintenance strategy

future modern

Founder, CEO
  • In partnership with Palantir data scientist Andrew Keller; CERN computational physicist Giordon Stark; Complex “25 under 25” artist Big Baby Gandhi; OWSLA co-founder, Skrillex’s former manager, and Audius Head of Partnerships Clayton Blaha; Bearpaw CEO Tom Romeo; and 700+ more partners
  • Agency and aspiring product studio offering creative, technical, and professional services specialized in blockchain and machine learning
  • Advise data science for client Tom Romeo’s AI lead generation product, targeted for a $90 million acquisition
  • Internally building Cloudchaser, which does unsupervised learning on the social graph to deliver the web’s first open-source, human-centered, spatial content discovery tool. Mined over 600K social media profiles into a graph database using Python, neo4j, and APIs for popular UGC networks like Soundcloud. Implemented a PageRank algorithm for web profiles, which produces the ten highest-ranked artists out of clusters of ~20,000 in ~10 minutes on a local machine.

Verifi

Product Analyst
  • Worked under Director of Analytics Howie Shack
  • Built dozens of automated reports to track product and operational data, often within an hour of assignment
  • Identified 1,000s of errant chargebacks occurring monthly at a high-profile client.
  • Developed a SQL time series forecast that predicted the caseload over the next three months within 5% error

Core Competencies

Languages:

English, Japanese (7 years of study, 4 on HL IB Japanese exam), Twitter, Python, SQL, Spark, PySpark, Spark SQL, Matlab, Mathematica, C, C++, Java, Excel, LateK, HTML, CSS, next.js, Typscript, Tailwind

Tools and Skills:

gradient boosted trees, random forests, logistic regression, linear regression, mathematical optimization, unsupervised learning,network science and analysis, exponential smoothing, time series forecasting, A/B testing esp. non-normal distributions, data visualization, geo-spatial modeling, analytics and visualization, REST APIs, data mining, data scraping, ETLs, Databricks, Delta, Parquet, S3, Tableau, Mode, Superset, DataDog, Agile, Lean, CI/CD

Relevant coursework:

Game Theory; Industrial Organization; Quantitative Marketing; Product Design for the Developing World; Entrepreneurial Development; Machine Learning; Computer Systems; Algorithmic Economics; Economics of Uncertainty and Information; Network Structure & Economics (University); IB Design Technology, IB Business and Management (High School)

Education, Activities, and Accolades

California Institute of Technology

BSc, Economics; Minor, Computer Science
GRE Score:
168 Verbal
164 Quantitative
5.5 Analytical

Troy High School

International Baccalaureate Diploma Recipient
GPA: 4.70
SAT Score: 2340
780 Math
800 Reading
760 Writing