A Multi-layered Mamdani Fuzzy Soft Expert System is planned to diagnose HIVAIDS. The Proposed DHA-MLMFES System is an Automatic diagnose of HIV-AIDS exploitation Multi-layer Mamdani Fuzzy Soft Expert System (DHA-ML-MFES) knowledgeable System, will classify the various stages of HIV-AIDS like No-HIV-AIDS, HIV Stage 1, 2, 3 and Aids. The Expert System has seven input variables at layer-I and four input variables at layer-II. At layer-I input, variables are headache, fatigue, aching muscles, sore throat, swollen lymph nodes, red rashes, fever that detects output condition of HIV-AIDS, infection, and different downside. The additional input variables at layer-II are PCR+, PCR+ P24+, P24+ Ab+ (ELISA), Ab+ Wb-inderminate that verify the output condition of HIV-AIDS like No-HIV-AIDS, HIV stage I, II, III or AIDS and alternative infection and downside. Results display investigation of the accuracy of the outcomes up to 94 percent of exploitation planned DHA- Multi-layer Mamdani Fuzzy Soft Expert Structure to set up the advanced HIV-AIDS technique with help of physician assumptions and compile it to the medical Lab of Roscommon University Hospital, Ireland.