Real-world negotiations are characterised by combinatorially complex negotiation spaces, tough deadlines, bounded agent rationality, very limited information about the opponents, and volatile negotiator preferences. Classical negotiation models fail to address most of these issues. This paper illustrates our practical negotiation agents which are empowered by an effective and efficient genetic algorithm to deal with complex, incomplete, and dynamic negotiation spaces arising in real-world applications. Initial experiment demonstrates that our genetically optimised adaptive negotiation agents outperform a theoretically optimal negotiation model when time pressure exists. Our research work opens the door to the development of responsive and adaptive negotiation agents for real-world applications.