I am working on a security-critical tool that uses ECDH to establish shared session keys. I want to reinforce this process by using a PQ-KEM algorithm like Kyber. Right now, I am thinking of achieving this by having two independent key exchanges (one with ECDH keys and one using the PQ-KEM) and then deriving the shared key by passing the two derived secrets through an HKDF. Is this a good approach or am I missing something critical?