A platform offers an environment where heterogenous buyers and sellers have one opportunity to meet a partner and form a match; no match is feasible outside the environment. We evaluate and compare the profit maximizing and the surplus maximizing platforms. There are three important features: buyers and sellers only observe (inside the environment) a noisy signal of their partner's type and do not observe their own noisy signal, the platform can only charge a fixed (but possibly different) access fee to buyers and sellers, and the platform controls the signals. The platform affects the matching environment through two channels: through the effect on the probability of being matched, and through the expected type of buyers and sellers willing to participate. We compute and provide microfoundations to the platform's optimal pricing rule for a broad set of matching environments, and show which matching equilibrium emerges under several model's primitives. Finally, we show the profit maximizing platform overprovides information to high-type buyers and sellers, and underprovides information for the rest compared to the benchmark case. Our platform can discriminate between-groups and within-groups.