My approach: $$\lim_{j \rightarrow \infty} \int_0^j \left(1+\frac{x}{j}\right)^j e^{-\pi x}dx = \lim_{j \rightarrow \infty} \int_0^{\infty} \left(1+\frac{x}{j}\right)^j e^{-\pi x}dx.$$ I'm not sure if I can do the above, but I need it to apply Lebesgue as follows:
I see that $(1+\frac{x}{j})^j \rightarrow e^x$ and $(1+\frac{x}{j})^j$ is monotonic increasing, so $e^x$ is a dominating function that is integrable. Hence I can apply Lebesgue:
\begin{align*} \lim_{j \rightarrow \infty} \int_0^{\infty} \left(1+\frac{x}{j}\right)^j e^{-\pi x} dx & = \int_0^{\infty} \lim_{j \rightarrow \infty} \left(1+\frac{x}{j}\right)^j e^{-\pi x}dx \\ &= \int_0^{\infty} e^x e^{-\pi x} dx= \int_0^{\infty} e^{(1-\pi) x} dx= - \frac{1}{1-\pi}. \end{align*} My colleagues all have a different result though and Wolfram Alpha exceeds computation time. I cannot spot a mistake.