Why CombDesk exists
A link list is an unfinished answer. Classic search was built when the scarce resource was the index; the abundant resource was your attention. Both have flipped. Sources are everywhere, and nobody has time to read forty of them.
CombDesk is an intelligence search engine. It searches, reads, compares, verifies, organises and then answers, with the sources kept in view so you can push back on any part of it.
Not another Google
Competing on general-purpose retrieval alone is a losing fight, and an uninteresting one. The interesting territory is the work that happens after retrieval: elimination, eligibility, comparison, contradiction. That is where a question like "which of these grants could we actually win" either gets answered or does not.
Federate first, index later
CombDesk does not need its own global crawl on day one. It federates existing search, index and data providers, applies its own retrieval and ranking layer, and puts CombDesk reasoning and interface above them. Over time more of the index, and the knowledge graph around it, becomes proprietary.
CombDepth is the signature
Most search products give you one speed. CombDesk lets you spend more effort on questions that deserve it. At Quick it is an instant answer box. At Deep, sub-questions are combed separately, checked against each other, and disagreements are reported rather than smoothed over.