i struggle with stationary points for this problem, minimize:
$\text{f}(\textbf{x})=x_{1}^{2}-x_{1}x_{2}+2x_{2}^{2}-2x_{1}+\mathrm{e}^{(x_{1}+x_{2})}$
I know that gradient is:
$\nabla f (\textbf{x)}=(2x_{1}-x_{2}-2+\mathrm{e}^{(x_{1}+x_{2})}, -x_{1}+4x_{2}+\mathrm{e}^{(x_{1}+x_{2})})^{T}$
And also if I want local extrema I set them equal to zero:
$2x_{1}-x_{2}-2+\mathrm{e}^{(x_{1}+x_{2})}=0$
$-x_{1}+4x_{2}+\mathrm{e}^{(x_{1}+x_{2})}=0$
And suddenly I stopped because I'm not sure how to solve this kind of system.
Thank you for your hint :)

Changing notations $(x_1\to x,x_2\to y)$ you want to solve $$f'_x=e^{x+y}+2 x-y-2=0\tag 1$$ $$f'_y=e^{x+y}-x+4 y=0\tag 2$$ So $$f'_x-f'_y=3 x-5 y-2=0\implies y=\frac{3 x-2}{5} $$ Replace in $(1)$ and obtain $$\frac{7}{5}x+e^{\frac{8 x-2}{5}}-\frac{8}{5}=0\tag 3$$ which could be very easily solved using Newton method starting with $x_0=0$ as shown below (notice that the derivative of this new function is always positive). $$\left( \begin{array}{cc} n & x_n \\ 0 & 0 \\ 1 & 0.3760062343 \\ 2 & 0.3313943919 \\ 3 & 0.3304499020 \\ 4 & 0.3304494983 \end{array} \right)$$
Sooner or later, you will learn that any equation which can write or rewrite $$A+Bx+C\log(D+Ex)=0$$ has real and/or complex solution(s) in terms of Lambert function (which is not an elementary function).
The Wikipedia page shows many examples of the kind of manipulations to be performed. In your case, the solution is given by$$x=\frac{8}{7}-\frac{5}{8} W\left(\frac{8 }{7}e^{10/7}\right)\qquad y=\frac{2}{7}-\frac{3}{8} W\left(\frac{8 }{7}e^{10/7}\right)$$
Using the expansion given in Wikipedia
$$W(x)={L_1}-{L_2}+\frac{{L_2}}{{L_1}}+\frac{({L_2}-2) {L_2}}{2 {L_1}^2}+\frac{{L_2} \left(2 {L_2}^2-9 {L_2}+6\right)}{6 {L_1}^3}+\cdots\tag4$$ where $L_1=\log(x)$ and $L_2=\log(L_1)$ you would find that $$W\left(\frac{8 }{7}e^{10/7}\right)\approx 1.3061$$ (the exact value should be $\approx 1.2999$) So $$x\approx 0.3266\implies y\approx -0.2041$$ while the exact solution would be $$x\approx 0.3304\implies y\approx -0.2017$$ For sure, adding extra terms to $(4)$ would give something closer.