This is more of a constructive proof if you want to know how to actually compute these polynomials. But I also really like the simplicity of the algebraic proof mentioned by another user here, just by studying the kernel of a linear morphism between two vector spaces of the same dimension, and would consider that one as the best answer.
You can consider the functions
$$\begin{cases}L_{0, k}(x) = \frac{x^{k}}{k!} (1-x)^{k+1} \\ L_{1, k}(x) = \frac{(x-1)^{k}}{k!} x^{k+1} \end{cases}$$
So, these satisfy that
$$\begin{cases}L_{0, k}^{(n)}(1) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \\ L_{0, k}^{(n)}(0) = \delta^n_k \hspace{1 cm} \forall \hspace{0.1 cm} n\leq k \end{cases} $$ $$\begin{cases} L_{1, k}^{(n)}(0) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \\ L_{1, k}^{(n)}(1) = \delta^n_k \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \end{cases}$$
where $\delta^n_k$ is a Kronecker delta ($\delta^n_k=1$ if $n=k$, and $\delta^n_k=0$ if $n\neq k$)
Then, start constructing backwards.
Given polynomials $\{L_{0, k}, L_{0, k-1},...,L_{0, N} \}$ and $\{L_{1, k}, L_{1, k-1},...,L_{1, N} \}$ (for some $1 \leq N \leq k$) of degree at most $2k+1$ such that
$$\begin{cases} L_{0, m}^{(n)}(0) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N \leq m \leq k \\ L_{0, m}^{(n)}(1) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N \leq m \leq k \end{cases}$$
$$\begin{cases} L_{1, m}^{(n)}(0) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N \leq m \leq k \\ L_{1, m}^{(n)}(1) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N \leq m \leq k \end{cases} $$
define
$$\begin{cases} G_{0, N-1}(x) = \frac{x^{N-1}}{(N-1)!} (1-x)^{k+1} \\ G_{1, N-1}(x) = \frac{(x-1)^{N-1}}{(N-1)!} x^{k+1} \end{cases}$$
and
$$\begin{cases} L_{0, N-1}(x) = G_{0, N-1}(x) - \sum_{n=N}^k G_{0, N-1}^{(n)}(0) \cdot L_{0, n}(x) \\ L_{1, N-1}(x) = G_{1, N-1}(x) - \sum_{n=N}^k G_{1, N-1}^{(n)}(1) \cdot L_{1, n}(x) \end{cases}$$
It is immediate that the new set of polynomials $\{L_{0, k}, L_{0, k-1},...,L_{0, N}, L_{0, N-1} \}$ and $\{L_{1, k}, L_{1, k-1},...,L_{1, N}, L_{1, N-1} \}$ satisfies
$$\begin{cases} L_{0, m}^{(n)}(0) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N-1 \leq m \leq k \\ L_{0, m}^{(n)}(1) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N-1 \leq m \leq k \end{cases}$$
$$\begin{cases} L_{1, m}^{(n)}(0) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N-1 \leq m \leq k \\ L_{1, m}^{(n)}(1) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n \leq k \hspace{1 cm} \forall \hspace{0.1 cm} N-1 \leq m \leq k \end{cases} $$
and have degree $2k+1$ at most.
You can iterate this process until you get $\{L_{0, k}, L_{0, k-1},...,L_{0, 1}, L_{0, 0} \}$ and $\{L_{1, k}, L_{1, k-1},...,L_{1, 1}, L_{1, 0} \}$ satisfying
$$\begin{cases} L_{0, m}^{(n)}(0) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n,m \leq k\\ L_{0, m}^{(n)}(1) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n,m \leq k \end{cases}$$
$$\begin{cases} L_{1, m}^{(n)}(0) = 0 \hspace{1 cm} \forall \hspace{0.1 cm} n,m \leq k \\ L_{1, m}^{(n)}(1) = \delta^n_m \hspace{1 cm} \forall \hspace{0.1 cm} n,m \leq k \end{cases} $$
Then doing linear combinations of these polynomials you can impose any derivative of any order at $0$ and $1$, defining
$$f(x) = \sum_{n=0}^k p^{(n)}(0)L_{0, n}(x) + \sum_{n=0}^k p^{(n)}(1)L_{1, n}(x)$$
where $\{p^{(n)}(0)\}_{n=0,...,k}$ and $\{p^{(n)}(1)\}_{n=0,...,k}$ are the given values you wanted.
To see the uniqueness, you can consider two polynomials $f$ and $g$ of degree at most $2k+1$, such that $f^{(n)}(0) = g^{(n)}(0)$ and $f^{(n)}(1) = g^{(n)}(1)$ for all $n\leq k$. Now if we consider $h=f-g$, we have that $h^{(n)}(0)=0$ and $h^{(n)}(1)=0$ for all $n \leq k$, and $deg(h) \leq 2k+1$. So we have to see that a polynomial $h$ satisfying these conditions has to be $0$.
Such a polynomial should be divisible by $x^{k+1}$ and by $(x-1)^{k+1}$, which are coprime (no common factors in their unique factorization), which implies that such a polynomial has to be divisible by $x^{k+1}.(x-1)^{k+1}$, and therefore the degree is at least $2k+2$ unless $h=0$.