Alexandre Salama

➼ Data leadership: leading the data component of high-impact projects within several teams (Supply, Host Incentives, Ukrainian refugee response, COVID-19 response, Airbnb for Work (B2B), EMEA Business Operations), with product teams (PMs, engineers, designers) and operational teams (sales, marketing)

➼ Data products: design of 3 incentive campaigns for 200,000+ Airbnb hosts to encourage them to open their calendar by using data to 1- target the right audience, 2- create customized goals/rewards for each host, 3- measure ROI

➼ Data Scientist Management: management of Airbnb's COVID-19 response data teams, namely a task force composed of 3 DS who created a dashboard used externally (Tweets from Airbnb CEO, Press Relations etc.) and internally to understand our product (300+ internal visitors to the dashboard)

➼ Data engineering and analytics: creation of data tables and strategic metrics used by Airbnb for our 2020 IPO, concerning Airbnb for Work's 7M+ users

➼ Experimentation: creation of over 20 AB tests using various statistical methods (stratification, propensity score matching, causal trees) to quantify the impact of our initiatives

➼ Machine learning: design, training, and production deployment of an XGBoost model that predicts the propensity of a session on Airbnb to be related to a business trip, using over 60 features, on +100M users, producing an AUC of 0.90

➼ Causal inference: analysis to measure the performance of companies using Airbnb using several quasi-experimental techniques (diff-in-diff, synthetic controls, Lasso/Ridge regressions)

➼ Research: design of a neuroscience research project, using "near infrared spectroscopy" technology to understand what happens in the brain of two subjects interacting on topics they agree/disagree on

➼ Data analysis: collaboration with technicians to collect data from 50 subjects and analysis of behavioral data (language, facial expressions)

➼ Academic writing: co-publication in Frontiers of Human Neuroscience (“Interpersonal Agreement and Disagreement During Face-to-Face Dialogue”)

➼ Financial modeling: analysis of the performance of a $10bn real estate project portfolio for internal or investor presentations, using financial models (net present value, internal rate of return)