Julius Monello Email & Phone Number

Berlin, Germany

Berlin, Germany

Berlin, Germany

Berlin Area, Germany

– Built model to suggest pre-drafted responses (PDRs) to customer service agents using chat. Suggested changes to the subject and grouping of PDRs based on clustering of chat messages, and worked with Engineering and Product to integrate the algorithm into a new and improved chat system.– Worked with stakeholders across several different organizations within Wayfair to integrate Data Science work within the Operations & Supply Chain space into a company-wide platform for order visibility. Identified gaps in data coming from different teams to help streamline all relevant order-to-delivery data.– Within this order visibility space, created an algorithm to predict, as of any given day, the probability that an order will ship by its must-ship-by-date (MSBD); this was used in order to promote a more proactive approach to supplier outreach.

Portland, Oregon, United States

Joined as part of the Information Management & Analytics team to ramp up machine learning capabilities within the team; left after a few short months to pursue a Masters in Artificial Intelligence due to some unfortunate timing.

Portland, Oregon Area

- Designed an algorithm to optimally deliver guaranteed online advertising campaigns so that non-guaranteed ad revenue is maximized and delivery is proportional to web traffic patterns, using a proportional-integral-derivative (PID) controller along with traffic forecasts. Wrote a series of jobs and created dashboards in Python to help project stakeholders monitor and debug the algorithm.- Improved existing algorithm that probabilistically matches billions of users across different devices seen on the AppNexus platform. Wrote and maintained Hive jobs to generate and store the necessary training data in HDFS.- Researched features that are good predictors of bid prices to improve algorithm that dynamically sets reserve prices on real-time-bidding (RTB) auctions. Third author of paper describing this algorithm that won Best Application Paper at the IEEE DSAA 2016 conference in Montreal.

Beaverton, Or

- Monitor, report on, and forecast future performance in order to optimize net cash returns for a variety of credit card and student lending portfolios. - Operate as part of a small team in charge of the launch of a new credit card product; duties include creating reports to monitor customer booking and activity, identifying and fixing back end data errors, and communicating results and milestones with the rest of the company. - Analyze large datasets from other companies in order to help valuate potential new portfolio acquisitions.- Have expertise in data mining and analysis using SQL Server, STATISTICA, and Microsoft Excel.

Portland, Or

Chosen by faculty to assist students with mathematics and physics coursework; workwith said professors to adjust course syllabi in order to improve student comprehen-sion.

Reed College, Portland, Or

Numerically modeled a stochastic mass transport model that exhibits phase transi-tions, utilizing parallel computing methods on a distributive cluster. Wrote code tosimulate the mass transport and subsequently analyze the phase transitions. Worked under Professor Darrell Schroeter.

Las Vegas, Nevada

Using standard 1" optical equipment, independently tested a new pressure mediumusing high-pressure experimental methods with Diamond Anvil Cells. Analyzed andcompared data trends against established pressure media to conclude that the newmedium performs better at higher pressures.

Master'S Degree, Artificial Intelligence (Machine Learning Specialization)

Dissertation: "Subsampling from Feature Embedding Clusters for Calibrated Classification" Other interesting stuff: - Generation of novel.

Bachelor Of Arts (B.A.), Physics

Senior Thesis: Two-Dimensional Stable Configurations of BuckyBalls Commended for Excellence in Scholarship (2009-2013) Ultimate Frisbee