From Job Description to Shortlist in Five Days

Day 1: Job Intake

The five-day clock starts with a job intake call. This is the only step where humans do most of the talking. The hiring manager describes the role, the must-haves, the nice-to-haves, the team context, the comp band, and the kinds of profiles that have worked or failed in the past. The recruiter captures it, the AI structures it, and both sides confirm the criteria together before the workflow moves downstream.

Most failed hires are traceable to a vague intake. Spending a real 45 minutes here — and getting alignment in writing — saves an enormous amount of misdirected sourcing later. AI can speed up everything downstream, but it can't fix a misaligned brief at the top.

Day 2: Talent Discovery

Day 2 is when AI discovery does the work that used to take a recruiter most of a week. Inputs from the structured intake — role, skills, seniority, geography, comp range, exclusion lists — feed into an AI search across internal and external candidate sources. The output is a ranked list of profiles with context: why they match, what to look for in outreach, where the risk signals are.

A typical ranked pool for a knowledge-worker role lands somewhere between 80 and 250 profiles. The recruiter reviews the top of the list (not the whole thing), removes anyone obviously off, and approves the working pool. The time spent reviewing is the recruiter's chance to add the judgement the AI can't — pattern-matching from previous searches, knowledge of which companies are good signals, awareness of recent moves in the market.

For more on how this layer compares with manual sourcing, see How AI Reduces Hiring Time by 50+ Hours Per Role.

Day 3: Outreach and Screening

Day 3 is outreach. Personalized, sequenced messaging goes out to the approved pool — drafted by AI, reviewed by the recruiter, sent under the recruiter's name. Responses come back through the day. Interested candidates are auto-qualified against the structured criteria with a few short questions. Not-interested or off-criteria responses are filed cleanly so the team has a real picture of market sentiment, not just inbox noise.

By end of day 3, the pool has narrowed naturally. A typical 150-profile pool produces roughly 25–45 interested replies and 12–22 candidates qualified to advance to an AI interview. The recruiter spends the time saved on the conversations that need a human voice — passive candidates with questions, edge cases, candidates the AI flagged as high-potential but unconventional.

Day 4: AI Interviews

Day 4 is the AI interview round. The AI Interview Agent runs structured, adaptive conversations with the qualified pool. Each interview is recorded, transcribed, and scored against the rubric set on Day 1. Candidates can take the interview when it suits them, which compresses what used to be a two-week scheduling cycle into a single day for most of the pool.

The output isn't just a score. It's a structured evaluation report per candidate — strengths, concerns, sample responses, comparative ranking against the pool — that the hiring manager can read in five minutes instead of watching a 45-minute video. For more on the interview format, see Best AI Interview Software (2026).

Day 5: Ranked Shortlist

Day 5 is delivery. The recruiter reviews the AI interview outputs alongside the candidate context, calibrates against the original brief, and produces the shortlist: typically 4–8 candidates, ranked, with the evaluation reports attached. The hiring manager gets the shortlist in their inbox the morning of day 5, ready for human-led second-round interviews.

Crucially, the shortlist isn't an "AI-approved" list. It's a recruiter-curated list informed by AI evaluation data. The recruiter has read every report, knows which candidates are stronger than their score suggests, knows which red flags to highlight, and is prepared to walk the hiring manager through the trade-offs. The AI does the heavy lifting; the recruiter does the curation and the briefing.

Final Decision Remains Human

Everything above day 5 is preparation. The hire/no-hire decision belongs to the hiring manager and the team, working from a much better starting position than they would have in a traditional process. They get to spend their time on the highest-leverage activity — assessing the top of the pool — rather than working through the bottom of it.

This is the philosophy that underpins the whole workflow. Speed is not the same as automation of the final call. AI removes the work that doesn't require judgement so humans can spend more of their attention on the work that does. The five-day timeline isn't about replacing decisions — it's about getting better decisions to be made sooner, with better information. See The AI Recruitment Playbook 2026 for how this workflow fits into the broader recruiting operating model.