I shall expand further on some of my ideas:
When the Hivemind starts play, it will be a small brain bug containing the mind, and be accompanied by a small pack of minions to protect it. The genes accessible to start with will be low, as will its store of biomass. It will thus need to use its minions to hunt prey for biomass and genetic samples. In order to then use those samples, it will then combines the genes into a 'strain'. This strain can then be grown. Living minions cannot be upgraded in any way. It is recommended that obsolete members are to be recycled for their biomass. However, growing new minions takes time. The Hivemind produces minions like a farm, not like a factory. Even in optimum situations it will take days to produce a new minion, in sub-optimal (as is likely at the start) it will be in the period of weeks. The Hivemind has the advantage of biomass being pretty ubiquitous and recyclable, as well as having no real upper limit on how many minions it grows at once, given infinite biomass. Thus, Hiveminds should go for a quantity other quality approach. Beware, minions require biomass to live. Running low can quickly cause the death of the horde.
When the Commander starts play, it's databanks will have been corrupted due to the effects of deactivation for millennia. Thus it will only have access to a few modules for construction, more modules will have to be acquired in the field. It will start alone, save for a small omnifactory (using nano-bots, 3D printing and servo arms) capable of building anything up to a human in one go, and anything larger in IKEA-style kits. The Commander itself is a near-indestructable combat unit, and is capable in construction, with manipulators and a small nano-bot spray to piece together kits or, in desperate cases, build from scratch. The commander is very much a jack-of-all trades, master of none. It works best when assisted by specialist units. Factories can output units quickly, but multiple rare materials are required for anything but the simplest of machines. The commander can piece together different modules to create a blueprint, and feed this to a factory with the required resources in order to create its army. The Commander has the advantage that it can upgrade existing units to keep from going obsolete, allowing it to modify units in the field to react to the situation at hand. Thus, commanders should go for a quality over quantity approach. Beware, units require energy to be active, and the required resources for anything beyond dumb bots are rare.