30 min chat w/ recruiter; approached on LinkedIn. Emphasis of experience + ML systems dev for their ad/revenue engine. No hiring manager interview (red flag #1). Tech screen on HackerRank, one stated to be algorithms, one ML (hackerrank has no real ML questions -- red flag #2). Generally, a company hiring a staff (many years, 8+ usually) engineer will not use hackerrank as a hard screen; it filters for those who've recently finished a data structures course. If anything, a code sample or just live coding is good enough to detect fakers. Anyway, I was intrigued at the opportunity to waste VC money (the comp discussed was well under market for the role though and if you think Giphy junior equity is worth anything I have a bridge to sell you). So, while watching GoT I did the first tech q -- nothing big here, basic coding question. I was told the second would be an ML question, but I was confronted by a BFS/Dijkstra style graph traversal (in essence, given a starting node, print all other nodes in order of distance, with ties broken by the node id -- use a priority queue with a custom priority). At this point, it was clear that Giphy's hiring process was as broken as their business model -- I don't expect a recruiter to know what an ML question is but their hiring manager certainly should have (and backpedaled when I told them that it wasn't what they said it would have been). Not worth continuing -- insofar as it ever was -- so I left the problem partially finished -- just like the GoT ending. Tl;dr -- you're gonna be judged on hackerrank questions. Just keep a few of the hackerrank solved examples around.