You'd spend years firing pellets in advance of the voyage, on the same trajectory but moving a little slower than the ship you plan to launch. Time it all so you catch up with the first of them just as you want to start decelerating.

But then the whole project takes thousands instead of hundreds of years

And also it requires extraordinary navigation precision. These launches are going to be spread in time, which means that they don't happen from the same place (all celestial bodies move relative to each other). Which in turn means that when you approach the pellet there is going to be perpendicular velocity component.

Even without taking this into account all systems are going to need active guidance - there is no way you do single burn and end up in the same place in the same time few light years away.

The best schema is probably launching all fuels containers that should be picked up for retro burn at the same time with much higher acceleration (orbital rail gun and then a burner), and with very slightly smaller final speed.

Then on the main craft you do short retro burn to match speed to fuels containers, attach them and do full retro burn for orbital insertion.

Benefit of this schema is that you don't need to accelerate everything slowly using main engine, so your original booster can be way smaller. (Energy-wise it only works if you have stationary means to accelerate high-g-capable payloads, i.e. orbital railgun, in this case you don't have to accelerate extra fuel required for initial burn)

I didn't say it would be easy :-)

But orion pulse nuclear propulsion is pretty powerful stuff, especially if you can antimatter-catalyze it.