This is an attempt at an answer to an older version of the question, where the focus lay only on cooling the milk as fast as possible.
This is only a partial answer - I have no idea whether my "method" is the fastest.
Heat transfer can be done in three ways: convection, conduction and radiation. We cannot really influence radiation, but it doesn't help much anyway (the temperatures are too small). Roughly, thermal conductivity is governed by Fourier's law, which tells us that the local heat flux $q$ is roughly
$$ q= k\nabla T$$
where $\nabla T$ is the temperature gradient and $k$ is a constant depending on the material where the flux occurs (thermal transmissivity).
Therefore, we can influence conduction by
a) bringing the milk into contact with a material with high thermal conductivity (such that conduction is accelerated)
b) cooling this contact material as far down as possible
c) making the contact surface as large as possible
The third is necessary, because we cannot really influence the thermal transmissivity of milk. Even if the material in contact with the milk has infinite thermal transmissivity, this would just mean that regions closest to that material would be cold immediately and regions far away would still be warm and cool according to the conductivity of milk. Of course, we can remedy this by introducing a current, therefore introducing convection (i.e.: we can stirr).
Based on a)-c) and the common sense assumption that you don't want to create a total mess and you don't want to contaminate the milk (albeit what that means can vary), here is a suggestion:
(Maybe) fastest method: Buy a baking tray (copper would be ideal, but I'm not sure these exist. It seems to be best from pure alluminium, if this is safe, but stainless steel seems fine - here is a list of thermal conductivities), put it in your icebox and when you want to cool your milk, get it out, pour the milk onto the baking tray and maybe stirr a little (however, if the baking tray is large enough, the surface area of the 500ml of milk should be enlarged enough) and then recollect the cooled milk in your bottle.
However, there is a problem: While the thermal conductivity of metal is pretty good, its thermal capacity is very low compared to water (and therefore milk). This means that while the temperature exchange will be mostly instantaneous, it might not be enough to cool the whole bottle. Here is a back-of-the-envelope calculation: Heat capacity of water is about four times that of stainless steel (link). Temperature of the ice box is maybe -15°C and the milk would maybe be at 90°C when you get it out of the bottle. Therefore, in order to cool it to 30°C, i.e. the gradient is 60° for the milk and 45° for steel. This would mean that you need more than five times the amount of milk in steel (i.e. 2.5-3kg), which seems too much for a baking tray - and this ignores the fact that the temperatures will not instantly equilibrate, since the flux depends on the gradient.
Thus, you'd maybe need to cool the steel - preferably with water, because this is available (supposedly) in larger cold quantities. One way seems to be to spray the baking tray with cold water from below, but this could create a mess. Another way would be to hold the lower side in a bath of cold water. Therefore, here is the second method:
Method with more cooling power: We use our baking tray as above, but we also have a second vessel (maybe a larger baking tray) with cold water and we hold the baking tray in the cold water and maybe even stirr the cold water.
This however might be very cumbersome. So here's another proposition:
Slower, but enough cooling power: Don't pour the milk into a bottle, put it into some vessel which is very narrow or somehow far away from cylindrical or round form such that the volume to surface ratio is large (maybe some flask or a small bag that you can close) and completely submerge the vessel in cold water. Wave it around a bit and it should be cold quite fast. With this method, you won't get the thermal gradient and conductivity of the baking tray (however, water is still way better at conducting heat than air - see the list linked above), but since you cool water with water, a few litres of cold water will definitely be sufficient to cool the milk without heating up too quickly.