This sounds pretty cool: bipartite perfect matching is solvable in polylog time in parallel with poly many processors. This was known if you allow randomness but has been a longstanding open problem deterministically.
Since today Bipartite Perfect Matching is in NC. The proof uses connections between coding theory and Hall's theorem. Presented at WACT 2026. Yay!!!