The Wikipedia article Bicomplex number
has the key to the answer. It states that
The general bicomplex number can be represented by the matrix ${w\: iz \choose iz \:w}$ which
has determinant $w^2+z^2$.
Use a different basis and the matrix becomes
${a\:\: b \choose -b \:a}$ which has
determinant $a^2+b^2$ which form a complex
algebra which is a ring because they are closed under addition and multiplication with basis
$\,\{I,J\}\,$ where
$$I^2=I,\,IJ=JI=J,\,J^2=-I.$$
One matrix representation is
$$ I={1\:0 \choose 0\:1},\quad
J={0\:1 \choose -1\:0} $$
and another is
$$ I={1\:0 \choose 0\:1},\quad
J={0\:i \choose i\:0} . $$
The Wikipedia article also states
The bicomplex numbers form a commutative algebra over $\,\mathbb{C}\,$ of dimension two, which is isomorphic to the direct sum of algebras
$\,\mathbb{C}\oplus\mathbb{C}.$
The isomorphism is similar to the case of
split-complex numbers
$$aI+bJ\;\;\leftrightarrow\;\;(a+b\,i,a-b\,i)$$
where addition and multiplication are defined
componentwise which implies that the group of
units is $\,\mathbb C^{\times}\times\mathbb
C^{\times}$ with all numbers of the form
$\,a\,I\pm a\,i\,J\,$ being the zero divisors.