The peaks correspond to 1386 and 1635 cm-1 are related to the f-MWCNTs unless your ceramic doesn't possess these two peaks. It is better to do a control Raman spectra before drawing the conclusion.
I would definitely run a uncoated ceramic (TiO2?) spectrum as well as one with the MW-CNT's on another substrate. The lack of a peak at ~2700 may indicate a lack of deposition or a decomposition during the process. Did you do a baseline correction on this spectrum?
All the comments above are right. You should do 2 references (one the ceramic w/o CNT) to remove the Features that Comes from it, and a pure CNT without the ceramic. If there is a shift or Change of the Feature between the CNT and CNT/ceramic maybe you can conclude something about the interaction of CNT with ceramic, otherwise is probably only a phisical mixture.
I have explained the Raman spectra of MWCNTs in almost all of my articles. The D band at 1343-1350 cm-1 represents the defects and disorders in the sp2 C-C network of CNTs while G band at 1570-1580 cm-1 is representative of pristine arrangement of carbon atoms in graphitic structure of CNTs, in other terms metallicity or degree of crystallinity of MWCNTs. The chatacteristic ratio of these two bands ID/IG is reperesentative of quality of CNTs after several procesing stages i.e. dispersion, functionalization, consolidation etc. you can get more info from my articles.