It's concerning that this is for public safety….
I'm happy to see people trying to save money and do things on their own, but setting up a public safety system wrong can have some dire consequences.
Getting advice is a good start, but you'll need to understand that you are going to get some harsh words or advice you may not agree with. Understand that there are some of us that do this stuff for a living and have seen bad systems set up. Often we get brought in after the fact to "fix" it.
I'll also say that since you didn't bring up using amateur radio gear, Cheap Chinese Radios, etc. you are off to a better start than many others.
A couple of things you'd need to look at:
1. Get your licensing sorted out first. A coordinator will find a pair for you, and help determine how much power to license for.
2. Mobile radios often will not stand up well in a repeater application since the duty cycle can be high. They often will not handle the heat. Turning down the RF power quite a bit can help, but that can create coverage issues.
3. Some pre-made repeater systems, specifically the BridgeCom and others, are just two mobile radios in a box with a controller and sometimes a duplexer. Fine for low duty cycle use, but not for a primary dispatch system.
4. Avoid the temptation to go with cheap mobile/notch type duplexers. They can have a lot of loss and don't always work well in high RF noise environments. As Kayn1n said, avoid them.
If I were going to do this, I'd put money into the antenna, feed line and duplexer systems. If you do those right, you can have a well perming system. Often people will leave the antenna/duplexer choices to the last when most of the budget is spent and hope for a miracle.
If this is a secondary channel that doesn't get much traffic, then a couple of mobile radios might work, but keeping the power low will be beneficial to long life. Also keeping long winded people off them will help. Setting up time out timers on the system would be wise.
Controllers would be necessary, as kayn1n said. You also need it to do CW ID, as the system will need to identify itself.
This isn't an impossible project, but it won't be cheap. You didn't mention your budget, so it's hard to tell, but hopefully you have a few grand set aside for this. If you were hoping to do this for a few hundred bucks, you might want to rethink.