This may be a basic question but I am having trouble figuring out the correct answer.  I am trying to interpolate the local coordinate charts between two points $p_1$, $p_2 \in \mathbb{R}^N$.  The data I have is the ambient intrinsic dimension of the manifold (dimension d), the two points $p_1, p_2$, and an orthonormal basis for the tangent space at each point.  I would like to find the coordinates $(f_1(x_1, x_2, \ldots, x_d),  f_2(x_1,x_2,\dots,x_d), \ldots, f_N(x_1,x_2,\ldots,x_d))$ that could fit these restrictions.  I would like for each $f_i$ to be a polynomial function in $(x_1,x_2,\ldots, x_d)$ that best fits the data.  There is obvious ambiguity in this problem because a choice for $(x_1, x_2, \ldots, x_d)$ has to be made for $p_1$ and $p_2$.  I am wondering if there is a solution to this problem despite this ambiguity.  If anyone knows of any references to for this type of problem that would also be appreciated.