Web based information searching is increasingly becoming an essential day-to-day activity. Currently search engines are the only dominant tools used for searching information over the Internet. Most search engines are powerful tools in finding information and cataloguing them, however they are not much efficient in meeting the personalised information needs of users. This paper presents research work done for personalising web search practice by customizing search engine results via an agent called Personalised Search Agent (PSA). In this paper, we will describe the architecture, design and implementation of a PSA system. Finally, we show several experimental results obtained by comparing the performance of the PSA against traditional search engines.