How density fluctuations in the early universe led to the diverse galaxy population we see today is one of the greatest open questions in astrophysics. A key process we need to understand is the reionization of the universe from a predominantly neutral to predominantly ionized hydrogen intergalactic medium, which was likely caused by ionizing photons from the first galaxies. Studying the first generations of galaxies requires space telescopes because of the redshifting of light. I propose here to characterize galaxies at cosmic dawn (at redshift z<~10, <~ 500 Myr after the Big Bang). My project is motivated by two key science objectives pivotal to the NASA Astrophysics Research Program `Cosmic Origins' theme: (1) How do galaxies reionize the universe? (2) How does the galaxy population evolve at z>~7? These questions address NASA's strategic objective in astrophysics to explore how the universe began and evolved. I will aim to answer these questions by extending work done in two large Hubble Space Telescope surveys: the Grism Lens-Amplified Survey from Space, and the Brightest of Reionizing Galaxies survey, for which the data is already in-hand. The analysis and modeling of Hubble data will be complemented by ground based follow-up spectroscopy to characterize the completeness and purity of our HST samples, including a large program with VLT KMOS, which will be completed in the first year of this proposed research, and for which I am leading the design, reduction and analysis. I will use rigorous statistical techniques to analyze the data, and theoretical models will be built using semi-analytical techniques in combination with computer models. In addition, I will build on the results from my investigations to design successful Cycle 1 James Webb Space Telescope observations. These will characterize in detail the high-redshift galaxies identified by our Hubble surveys.