We analyze the general mathematical problem of global reconstruction of a function with least possible errors, based on partial information such as n terms of a Taylor series at a point, say the origin, possibly also with coefficients of finite precision. We refer to this as the "inverse approximation theory" problem, because we seek to reconstruct a function from a given approximation, rather than constructing an approximation for a given function.Within the class of functions analytic on a common Riemann surface Ω, bounded on Ω, or with the same rate of growth in the natural metric on Ω, and a common Maclaurin series, we prove an optimality result on their reconstruction at other points on Ω, and provide a method to attain it. The procedure uses the uniformization theorem, and the optimal reconstruction errors depend only on the distance to the origin.We provide explicit uniformization maps for some Riemann surfaces Ω of interest in applications. Some of these can also be obtained as a rapidly convergent limit of compositions of elementary maps. One such map is the covering of C \ Z by curves with fixed origin, modulo homotopies, precisely the one needed in the analysis of the Borel plane of the tritronquée solutions to the Painlevé equations P I -P V . As an application we show that this uniformization map leads to dramatic improvement in the extrapolation of the P I tritronquée solution throughout its domain of analyticity and also into the pole sector.Given further information about the function, such as is available for the ubiquitous class of resurgent functions, significantly better approximations are possible and we construct them. In particular, any chosen one of their singularities can be eliminated by specific linear operators which we introduce, and the local structure at the chosen singularity can be obtained in fine detail. These operators involve convolutions, whose singularity nature we analyze. More generally, for functions of reasonable complexity, based on the nth order truncates alone we propose new efficient tools which are convergent as n → ∞, and which provide near-optimal approximations of functions globally, as well as in their most interesting regions, near singularities or natural boundaries.