Three weeks from application to offer. Two of those weeks are calendar tetris. The other week is decision-making. Async screening attacks the wrong week first, which is why most async rollouts fail.
The Time-to-Hire Math
For most engineering hires, the longest stretch is between application and first technical screen — typically 8 to 14 days. Almost all of that is scheduling: waiting for the candidate's availability to overlap with an engineer's. Compress that to zero and you cut the funnel in half before you've changed anything else.
What Async Actually Fixes
Async fixes the scheduling tax, not the decision tax. If your hiring managers take a week to decide on a shortlist, async won't help. But if your bottleneck is the calendar — and for most teams it is — async first-round screening with an AI interviewer like Talia AI gives you most of the funnel compression for free.
The Async Stack
ATS → AI interviewer (round one, async) → engineer (round two, sync) → onsite. The async piece is the first round. Everything after is human, sync, and high-stakes — exactly where humans add the most value.
Watch-Outs
Two things kill async rollouts: completion rate (candidates don't finish) and rubric drift (scorecards aren't comparable). A real AI interviewer fixes both — completion rates are usually 80%+ because the interview is short, and the rubric is identical for every candidate. Pair this with the engineer-protection playbook and you have a complete operating model.
The Screening Room is an UpStack publication. Visit talia.ai.