Amit Misra Email & Phone Number

Redmond, Washington, Us

San Francisco, California, Us

I'm currently the lead for our resale optimization team. I build E2E models that help us determine our pricing policy for selling homes. These models drive key pricing decisions across the company, from setting our list prices, accepting offers, and projections for accounting.Technical Highlights*Developed and deployed our resale optimization model that manages thousands of homes and billions of dollars of inventory, driving $200MM in incremental quarterly revenue.*Built out causal forecasting tools to understand how our key business metrics would respond to changes in our pricing policy. *Generated key outputs used for human-machine interaction in pricing decision makingLeadership Highlights*Conducted 100's of interviews from junior ICs to Senior Managers at the company*Increased recruiting efficiency by 40% through better interviews and better interview calibration*Mentored teammates across technical and non-technical skillsets ranging from fresh grads to senior scientists and analysts

Redmond, Washington, Us

• Manage and coach a team of mid-level and junior applied scientists• Implement reinforcement learning for marketing at scale across Windows and Office for over 1B+ impressions/day • Build, validate, deploy, and track machine learning models in Databricks that scale to 100M+ users• Develop the machine learning roadmap for consumer marketing and drive to completion in collaboration with partners across engineering, marketing, product teams and finance

Redmond, Washington, Us

• Designed and conducted causal inference analyses (e.g., diff-in-diff, synthetic controls, regression-based approaches) to help shape $100M+ investments and initiatives • Proposed and evaluated A/B Experiments, generating $20M/yr in revenue• Built propensity models (xgboost, random forest, logistic regression, etc.) for targeted marketing that increased campaign impact by more than 4x

Redmond, Washington, Us

Redmond, Washington, Us

Implemented a metric to measure interdisciplinarity in journal citation networks. Tested metric, applied it to a 10,000-node network database and compared results over time spans of more than a decade. Presented results in an oral seminar to an interdisciplinary audience and in a research paper to be submitted.•Analyzed 10,000-node network database•Applied advanced graph visualization programs•Applied statistical tests to determine accuracy of metric

PhD Student in the Astronomy department working with Vikki Meadows in the Virtual Planetary Laboratory. Developed models for simulating exoplanet spectra and validated models against solar system observations. Made predictions for observational features that could be detectable with future ground and space-based telescopes.•Led a team of 3 Ph.D. students on a self-initiated research project*Developed, tested and documented 1,000-line Monte Carlo multiple scattering module to model exoplanet spectra*Created and maintained a 14,000 object database of model predictions*Validated model against existing Earth observation data to within 6%*Cleaned and processed 5.7 Gb of Earth observing data as model inputs*Developed new method to classify exoplanet atmospheres that improved on previous methods by a factor of 4*Discovered new techniques to characterize exoplanet clouds, volcanism and pressure and demonstrated their applicability to current and future telescopes

•Taught course on planets for adults over 50 from a wide variety of backgrounds•Developed original course material

Doctor Of Philosophy (Ph.D.), Astronomy And Astrobiology

Master'S Degree, Astronomy

B.S. Physics W/ Math Concentration, Physics, Mathematics, Astronomy