Loading…
Loading…
Every application scored against the role’s criteria the same way, with the reason next to the score; passive candidates messaged about their actual work, not a template; interviews confirmed and rescheduled without the calendar tennis. Three blueprints, in the ATS you already use.
Repeating work, per week17 h
Three blueprints cover most of it. The rest stays yours.
Three rungs
One-time prices, no subscription. Every rung is credited in full against the next one for 30 days, so starting small costs nothing extra. Card checkout on Stripe; the confirmation and the next step arrive within a minute.
$49one-time
You have imported a workflow before, or want to try before anyone touches your accounts.
$249one-time
You want the first blueprint running in your accounts this week, without learning a tool.
$490one-time
You want the whole job automated — all three blueprints, in your voice, with someone to call.
Every rung is credited in full against the next one for 30 days: kit → install → package. Model usage and any paid n8n, Make or Zapier plan are billed by the providers to your own accounts, usually a few dollars a month. Need more than the package? When a recruiter needs a custom build. Package terms: fixes, credits, refunds, your data.
What gets installed
Each is a documented blueprint from the library — trigger, steps, guardrails, free template — adapted here to how recruiters work. The kit and the install cover the first; the package covers all three.
Scores every application against the role’s criteria the same way, explains each score in a sentence, and hands you a shortlist you can disagree with.
Researches each passive candidate and writes a message that references their real work, then follows up twice inside sending limits — and stops the moment they reply.
Offers slots from the panel’s calendars, confirms by email or SMS, handles reschedules in the same thread, and cuts no-shows.
What it actually does
Not a feature list — the messages and actions as they happen once the first blueprint is installed. Every outgoing message starts as a draft you approved once; after that it runs on the rules you set.
Worked example · illustrative
Not a client report
About 340 applications a month across four roles. Eight hours a week reading them, usually in two long sittings; the shortlist depended on which recruiter read the pile, and hiring managers waited four days for it.
The first month| Metric | Before | After a month |
|---|---|---|
| Hours a week reading applications | 8 | 1.5 |
| Time from application to shortlist | 4 days | next morning |
| Scores overridden by the recruiter | — | 3%, logged |
| Applications scored the same way | no | all 312 |
This is a worked example, not a client's results: the figures are what the blueprint's own metrics and the role's published benchmarks make typical for an install of this size. Real numbers will replace it, with permission, once the first installs have run a month.
Which rung
Your tools, your volume, how much you want to touch. The answer is a rung and a price — no call at any rung. If you would rather write three sentences, the form is below.
Not on the list? Most tools with an API work; say which on the form.
Answer all three and the rung appears here — with what happens next and the price.
What happens after you pay
Card checkout on Stripe, then the confirmation lands within a minute. This is what it says — the same for every role, with the tools and blueprints swapped for yours.
Before you order
The screen applies the criteria you write down, the same way to everyone, and shows its reasons — which is more consistent than a tired reader at 6 pm. It shortlists; it never rejects on its own. If you hire into New York City or the EU, automated-decision rules apply and that version is a custom build with the audit and notice features they require.
The message references something specific they did, is sent from your address, and stops the moment they reply — from there it is you. Candidates notice generic messages, not well-researched ones.
Most ATS features summarise a resume. This scores against your criteria, explains the score, writes outreach and runs scheduling — three jobs, in your ATS, and you own the rules.
Outreach runs by email from your own address. LinkedIn steps are possible but carry their own limits and tooling, so they are part of a custom build rather than the package.
Start with the install — screening only, $249. The package makes sense once you are running three or more roles at a time, and the install is credited against it for 30 days.
When a package isn’t enough
A custom build starts at $2,000: your own rules, delivered into your accounts to keep, with documentation and 30 days of fixes. The package is credited against it for 30 days.
How custom builds workAn agency with several recruiters sharing a pipeline, with assignment rules and a manager view.
Hiring into a jurisdiction with automated-decision rules — NYC Local Law 144, the EU AI Act — with audit and notice features.
A client-facing portal where hiring managers see shortlists and book interviews themselves.
Order, or ask
One reply within 24 hours: the rung that fits, the payment link, and the next step. Everything happens by email and in your accounts — there is no call at any rung.
Prefer email? ask@agentmelt.com
Other packages
Speed-to-lead · Long-term nurture · Listing descriptions and posts
From $49 →Executive assistantsInbox triage · Meeting briefs · Weekly report
From $49 →Legal professionalsContract review · Document drafting · Client intake and inbox
From $49 →AccountantsInvoice processing · Daily reconciliation · Client reporting
From $49 →Accounts payable teamsInvoice processing · Purchase orders · Payment reconciliation
From $49 →Support teamsTicket deflection · Proactive outreach · Feedback analysis
From $49 →HR teamsNew-hire onboarding · Offboarding · Interview scheduling
From $49 →