🎤 New · Phase 4 shipped

Stop reading resumes. Hear them handle your objections.

Every HireSwiftlee applicant completes a short AI roleplay generated from your campaign — your customer's pain, your top objections, your tone. You scroll through audio clips and per-rubric scores in an hour, not a week.

Free for every client · Roleplay generated automatically from your job description · Available on every Tier 1-4 placement

📊 Your roleplay comparison · 32 candidates
#1
Maria Hernandez
Monterrey · ▶ 2:14 audio
87
#2
Carlos Rivera
Tijuana · ▶ 2:14 audio
81
#3
Ana Lopez
Guadalajara · ▶ 2:14 audio
78
#4
Jorge Castillo
Mexico City · ▶ 2:14 audio
71
+ 28 more · ranked by best of 3 attempts
Don't take our word for it

Take a sample roleplay. Right here. No signup.

5 minutes. Same engine your applicants use. Get scored at the end — see for yourself how rigorous it is.

🎤 Try it yourself · 5 minutes · no signup

Pick a scenario and take the roleplay — same engine your applicants use.

Nothing is saved or sent anywhere. This is for you to feel the experience your applicants will get.

The honest version

Why "they're pre-vetted" doesn't actually help you.

Toptal, Andela, Turing — they all sell the same pitch: trust us, they passed our test. The problem: their test is generic. It can't tell you whether this specific contractor is good at the specific work you need done.

Generic vetting tells you nothing role-specific.

"Passed our English test" doesn't tell you if they can handle a Texas property manager pushing back on price during a roofing cold call. Different job, different skill, no signal.

You're screening with resumes anyway.

So you read 60 LinkedIns, schedule 8 interviews, do 6 of them, hire 1. That's a week of your time and you still don't know how the hire handles your real objections until they're on payroll.

Marketplaces optimize for THEIR margin, not your hires.

Toptal's 3-5x markup pays for their static vetting. You pay for it whether the vetting was useful or not.

And the contractors hate it too.

Generic tests punish real talent who don't happen to perform well on artificial English exams. The contractors who would crush YOUR role get filtered out by tests that aren't about your role.

How it works

Three steps. About 90 minutes of your time total.

From posting the role to picking your candidate. Yes, really.

01

Post your role (5 min)

Title, description, top objections you'd expect a candidate to face. We generate the AI persona, customer context, scoring rubric, and opening line from your description — automatically.

02

Contractors complete the roleplay (passive)

Every applicant runs through your campaign's AI roleplay before their application reaches you. 5-7 minutes per attempt; 3 attempts max. Audio + transcript + LLM-scored.

03

You scroll 30 candidates in 45 min

Open the comparison page. Ranked by best score. Hit ▶ on the top 10 to hear how they handled YOUR objections. Expand summary on the 5 standouts. Schedule 2 interviews.

What's in the score

Five rubric criteria. One overall score. Plain-language summary.

The LLM scorer is calibrated against an anti-grade-inflation prompt. 90+ = exceptional. 75-89 = definitely interview. 60-74 = mixed. Below 45 = significant problems. The summary is written FOR YOU — what handled, what missed, red flags, standout strengths.

Sample rubric · Outbound SDR
Objection handling30%
Discovery questions20%
Tone + warmth20%
Listening + adaptation15%
Closing + next steps15%

Rubric weights generated automatically from your campaign. You can edit them later.

Compared head-to-head

HireSwiftlee vs. Toptal vs. Andela vs. Upwork.

Where role-specific evidence-based screening sits in the LATAM / nearshore talent landscape.

Capability
HireSwiftlee
Toptal
Andela
Upwork
Per-role screening
✓ Generated from your campaign
Generic English test
Generic English test
Self-reported skills only
Hear them handle YOUR objections
✓ Audio + transcript per candidate
✗ Resume only
✗ Resume only
✗ Resume only
LLM-scored rubric
✓ Per-criterion + summary
Compare 30 candidates in 45 min
✓ Side-by-side comparison page
Auto-shortlist top scorers
✓ Configurable threshold
Improvement curve (3 attempts)
✓ Coachability signal
✗ One-shot only
✗ One-shot only
✗ No screen
Bilingual ES/EN by default
✓ AI persona speaks both
✗ English only
✗ English only
✓ Some
Cost to client for screening
Free
$$$ rolled into 3-5x rate
$$ rolled into 2x rate
You build it yourself

Capabilities reflect publicly documented platform features as of 2026-06-03. If you're evaluating, talk to all of us — the differences become obvious in 20 minutes.

Pricing

Free for clients. Always.

We don't charge a screening fee. The roleplay is what makes our flat placement fee worth it — your hires actually pan out, so you stay, so we get paid more times. Aligned incentives.

What you actually pay

Hourly rate + flat 15% platform fee on every contractor invoice. That's it. The roleplay screen, the comparison page, the LLM scoring, the audio storage — bundled.

One-time placement bonus on first invoice ($200-$800 by tier) covers our marketplace + payroll work. We invest that into making the screen better.

See full pricing →
Roleplay screenFree
Audio storage + playbackFree
LLM scoring + rubricFree
Side-by-side comparisonFree
Auto-shortlist thresholdFree
Per-campaign scenario generationFree
FAQ

The questions we get on every sales call.

Won't contractors cheat by reading scripts from the screen?

The AI persona is generated fresh per campaign and varies its objections per attempt — there's no fixed script to cheat from. And cheating doesn't help in the rubric where it matters: the scorer weights tone, listening, adaptation, and natural objection handling. A canned-script reading scores like a robot.

What about contractors who are great but bad on the spot?

Three attempts, all visible to you, scores retained. The improvement curve is itself a signal — going 62 → 71 → 79 is coachability data you can't get any other way. A candidate who's consistently 78 across three attempts is probably a better hire than one who lucks into a 92 once.

Can I edit the AI persona / objections / rubric?

Yes. The scenario is auto-generated from your campaign description but you can edit any field — persona description, customer context, top objections, opening line, rubric weights. Regenerate with one click if your role pivots.

How accurate is the LLM scoring really?

Calibrated against an anti-grade-inflation prompt. We've tested it on roleplays our AM team also scored manually; agreement is in the high 80s%. We surface the LLM's reasoning so you can spot-check — every score links to a transcript moment that justified it.

Does this work for bilingual ES/EN roles?

Yes — and this is where we genuinely beat Toptal. The AI persona auto-detects that a CSR role is bilingual and surfaces Spanish during the roleplay (a Mexican customer would naturally code-switch). Browser TTS + STT support both languages. Whisper STT upgrade (Phase 3) makes accent handling even better.

What about TCPA / regulatory issues with AI in hiring?

AI roleplay is a skills evaluation, not an AI-mediated hiring decision. The hiring decision is still yours, and you have full transcript + audio to substantiate it. We don't use AI to make automatic accept/reject decisions on candidates — only to auto-shortlist when you opt in (defaults to off if you want full manual review).

Can I see a demo?

Post a free role and take the roleplay yourself — same experience your contractors will get. Or email sales@hireswiftlee.com for a 15-min walkthrough of the client view.

Replace a week of screening with a 45-minute coffee.

Post your first role free. Roleplay generated within 60 seconds of campaign creation. First applications start landing in 24 hours.

Post a role freeTalk to sales