Positioning with ints instead of floats

I round my transform.position to the nearest integer at the end of every frame so it’s never a decimal number. I would like to know if there’s a way to make unity store the position in ints instead of floats to prevent floating points precision problems.

No, it is not possible. Also, if you round to ints you will still have pretty good precision, because you simply do not store with decimal points.

To avoid floating point errors completely, you will have to move all objects by am offset to make a character go back closer to the zero coordinates.

I do not believe this is possible. You can store the values as int but Unity will always use decimals for the actual positional representations, so at the point the int information is handed back to Unity, floating point error is reintroduced. The error will be very small. If you are relying on int information you need to write a simple function to truncate or round your floats when you attempt to use them as ints. This totally negates the need to store as int; however storing as int may serve to reduce error, marginally.