You should know that "mathematically unbreakable because it is based on one-time pad" is a huge red flag. If their algorithm "combines one-time pad with other things" then it is not "mathematically unbreakable." If they don't know this, then you are absolutely dealing with a crackpot. If it is mathematically unbreakable then it is also certainly useless, because unbreakable encryption is known to be inherently very limited. If it is not mathematically unbreakable, then in my opinion it is highly unlikely to be better than standard encryption developed through global, open standardization efforts.
Finally, in my opinion, new encryption algorithms are simply not viable for commercialization. The cryptography community builds confidence in security by open scrutiny of algorithms, not secrecy. The community will never widely adopt new cryptography that is encumbered by patents or intellectual property restrictions -- there is precedent for this (see OCB, which is arguably better than more popular AEAD schemes but was never widely adopted due to its patents, even in light of a very generous blanket license). As I mentioned above, it is highly unlikely that this new encryption has any actual competitive advantage over the state of the art. Even if it did, no one is willing to pay for an encryption algorithm.