A square matrix is unitary if its transpose is its inverse i.e. \(U^{-1} = U^T\). One key utility of unitary matrices is that they leave the dot product invariant i.e.
\[\langle U x, U y \rangle = \langle U{-1} U x, y \rangle = \langle U^T U x, y \rangle = \langle x, y \rangle\]