Following the global economic downturn and a collapse in international equity markets, many financial institutions and corporations have sought the higher returns associated with higher risk from trading in foreign exchange derivatives. These derivatives have become increasingly complex to the point where few specialists are able to accurately determine the level of exposure. Top traders seek high rewards for their successful investments. Rogue traders seek high rewards by concealing their unsuccessful gambling, sometimes to the extent of endangering the viability of their employers. Current technology copes poorly with dynamically changing business requirements and conditions so there is little technological support available for organizations sensitized by reports of rogue trading and increasingly obliged by financial regulators to improve their risk management practices. This paper proposes a risk management framework that can support FX derivative monitoring and trading based on the Williams-Elliot Agent Architecture. The framework uses agent technologies for improved management of treasury risk by continuous monitoring of all transactions across an organisation; continuous evaluation of exposures compared with prescribed parameters across an organisation; instantaneous reporting to senior management where trading begins to approach or violates the parameters. Rigorous examples of typical transactions illustrate how intelligent agents can be used to monitor risk and to make trades within a powerful risk modelling and management framework.