Offshore
In an offshore model the work is carried out by a team at the supplier's engineering centre. It is the most cost-efficient model, and it suits high-volume and routine engineering activities with clear inputs and deliverables.
It depends on a good written brief, regular review calls and an agreed checking method, because the team is not sitting beside yours.
Onsite
In an onsite model the supplier's engineers work beside your team. This helps the supplier establish best practices in your organisation, absorb your product knowledge and share processes and technology with the offshore team.
It costs more per engineer than offshore, so it is usually reserved for work that needs close collaboration, access to physical products or plant, or fast learning of your environment.
Hybrid
A hybrid model puts a small onsite group in front of a larger offshore team. The onsite engineers take requirements, answer questions and review results. The offshore team carries out the bulk of the work.
It suits work of medium to high complexity, where coordination and integration with your systems matter but the full volume of work does not need to be done at your site.
How to choose
Four questions usually decide it. How complex is the work? How much of it is routine? How much knowledge must be transferred from your team? How much access does the work need to physical products or plant?
Routine, high-volume work points to offshore. Knowledge-heavy or access-heavy work points to onsite. Complex work with a lot of routine content points to hybrid.
Changing the model over time
A common pattern is to begin with a small offshore pilot, add an onsite lead as the scope grows, and move to a hybrid structure. Others start onsite to transfer knowledge and move volume work offshore later.
Whichever way you start, agree early how intellectual property, data access and knowledge transfer will be handled. If you are weighing individual engineers against a managed package, read augmentation or a defined work package.