It's not by any means my specialty but on what kind of time scale are you thinking of and how particular do you mean? One of the features of the outbreak of World War I is actually how slowly it occurred. Franz Ferdinand's assassination, which is typically attributed as the primary "cause" of the war happens on 28 June, is then followed by the July Crisis, and then full war between the belligerents doesn't actually break out until early August.

That some people thought crisis between the death of Franz Ferdinand and the outbreak of hostilities might result in conflict is somewhat obvious. That the result would be "World War I", with its horrific and interminable casualties, was less obvious.

So within that division it's not just necessarily a question of who thought some kind of war might be coming, but who thought some kind of incredibly destructive war might be coming. From what I've read, among the latter category we might include Kitchener and I believe Falkenhayn.

That some people had thought a war was possible or predictable seems clear through the actions in the various crises even prior to the assassination of Franz Ferdinand and the outbreak of WWI, so for instance both Austria-Hungary and Russia mobilize troops during the Balkans crisis. We could say the same, more or less, for crises like Agadir or Fashoda as well which everyone understood could have resulted in war, even if their relative diplomatic positions made that unlikely.

As a separate category you also had sort of long term thinkers like Jan (Ivan) Bloch who were writing about how industrial war between great powers in the 20th century would be catastrophic. One of the problems with Ivan Bloch though is that he's right for basically the wrong reasons, so some of the military scientific claims he makes in "predicting" what a World War look like are just wrong (e.g. speed and size of bullets, if I'm remembering right). So even though his analytical assessment is correct, military thinkers look at these technical points and know that he's wrong about why.

So I guess it really depends on what you mean.