Wow. You have multiple problems.
A '12V' lead-acid battery can have 2.2V/cell (or more) = 13.2V at the end of a recharge cycle. 13.2 x 28 = 370Vdc, so you're over the maximum input. Whilst the freshly charged battery terminal voltage will drop quickly upon load, it will be >357V for a non-trivial period of time. I would drop back to perhaps 25 batteries.
It's a 40 HP motor & VFD, which will happily draw 100A DC at full power, and a few times that under more demanding circumstances (rapid acceleration, locked rotor, etc, as noted in the motor's datasheet). I note in a comment you say you only need 10 HP, so that's good - but ensure you program the VFD to be deliver no more than that 10 HP / 7.5kW, because 40 HP / 30kW = ~100A from a string of 81 AH batteries is not how you want to treat 81 AH batteries, at least not regularly. And several times that, which is multiples of your C-rating, would be quite unwise.
As for your EE's plan to make a bigarse box of relays to reconfigure them from 28 in series to 28 in parallel, that's... cute, but unwise, because:
(a) relays that will carry several 10s of Amps continuous aren't cheap
(b) that's a wiring nightmare of materials & cost inefficiency (you're engineers, not scientists!)
(c) 28 batteries aren't all going to age the same way over time, they will discharge to differing V levels (growing internal resistance), then when you slam them all in parallel they're all going to try to equalise with uncontrolled currents flowing between them, which is not good battery treatment policy
(d) if 1 battery fails prematurely for whatever reason, and gets replaced, it won't have the same discharge characteristics, it will likely have a lower internal resistance, which will manifest the (c) problem even more - the new battery will tend to dump even more power into its peers
(e) that uncontrolled current has to be handled by the relays, which dramatically increases their current-handling spec & cost. Note: DC relays of X Amps rating are much more expensive than AC relays of the same current rating (it's to do with DC arcing during switching events and the much greater mechanical complexity of quenching that arc, whereas AC has the benefit of crossing 0-amps frequently
(f) recharging 28 x 12V 81AH batteries in parallel over, say, 20 hours (i.e. the typical recommended C/20 recharge rate = 4 Amps), requires a charger capable of ~120 Amps, which is yet more thick wire and another expensive (battery recharger) component.
Lastly, high voltage DC is dangerous, way more than AC at the same voltage. If you touch it, it doesn't let go - your muscles clamp on to the live conductor. This is quite unlike AC where the 50/60Hz reversal gives you some opportunity to let go consciously. Arcing in the event of short circuits poses critical danger to those working on it. I'm talking gloves and perspex face-masks, and TRAINING. The fact your EEs even suggested this relay box idea tells me they're not trained for this, at all.
The problem of course is that recharging a bunch of batteries in series is no walk in the park, either! The need to keep ALL of the batteries at very close to the same stage-of-charge is very real. Think about it, if 1, or some, of them falls behind and reaches flat first, then you have a choice to make (a battery management policy to decide): stop the whole system and start a recharge and hope they all reach 100%, or let the prematurely low SoC battery(s) become critically over-discharged but get the full discharge capacity from the battery-bank. There's various ways to keep a bunch of batteries in series at the same SoC, but that's a different question :-), albeit with no simple or cheap answer.