AI sales assistant. Reads, scores, writes the reply.
Every inbound lead is scored and answered in your own language. The draft is waiting before your rep opens the thread, and the rep is the one who sends it.
What it does
Inbound qualification
Every form fill lands as one record and is scored against the three questions that decide fit.
Draft-first replies
The answer is written in your approved language before a rep opens the lead, and it waits for them to send it.
Handoff with context
The rep gets the draft, the score and the reason in one place, so nothing is retyped and nothing is guessed at.
Nothing invented
A price or a promise the assistant was never given comes back marked as pending you, with the gap named.
How it works
Capture
A lead fills the form and the record lands with an email to you.
Qualify
The lead is scored against your ICP and your three fit questions.
Draft
The first reply is written and waits in the queue, price claims flagged.
Send
A rep reads the draft, sends it, and takes the conversation from there.
0
Messages that leave without a person on the send
24/7
Inbound form leads accepted, weekends included
+20–40%
Meetings booked
Target range we model in pilots
15–20h
SDR hours handed off weekly
Target range we model in pilots
What the first week changes
The wait between a lead arriving and an answer being ready collapses: the draft is written before a rep opens the thread. A 20–40% lift in booked meetings is the target range we model in pilots, before any rep changes behaviour.
Every run is written to the journal with its inputs and its outcome, failures included, so a week of work can be read back rather than reconstructed from memory.
What a deployment looks like
The AI SDR covers the hours no one is at a desk, so the human SDR spends the day talking to the right prospects.
Modeled scenario: a 22-person B2B software company puts an AI SDR on inbound. Within a month the human SDR shifts from chasing leads to running discovery calls and closing. The tiles are targets we scope a pilot against, built from four inputs we show in full on the scenario page.
+34%
Meetings booked
4h → minutes
Time to first reply
87%
First replies that start from a draft
18h
Human SDR time freed weekly
Targets for a pilot of this shape · see the arithmetic
Evidence · AI SDR
What runs today
The parts of this role that exist in the platform right now, and the parts a deployment has to build with you. Nothing in between.
Ships today
Any form on your site can post a lead into one record
The endpoint runs in production. It caps the request body before parsing it, absorbs bot submissions on a hidden field without writing anything, and stores the lead in Postgres. A submission that fails validation gets a reason back instead of a silent success, and the form on your own site also emails you the moment a lead lands.
Ships today
The reply is written ahead of the rep, and the rep sends it
Speed to first contact comes from having the answer ready, so the draft is composed before anyone opens the lead. The role has no send path of its own. The phone number, the email address and the full name stay out of the prompt — it works from an initial and the answers on the form.
Scoped with you
CRM and calendar start as a scoping conversation
One platform connector ships today, and it is Meta ads. A CRM, a calendar or a sequencer is set up against your systems at the start of a deployment. We name the connectors we have built and leave the logo wall to someone else.
The gates behind the drafts — approvals, spend caps, the kill switch — are listed on the product page.
Frequently asked questions
No. It handles the top of the funnel so human reps spend more time on calls and negotiations.
You brief it in plain language. It scores leads against your rules and escalates edge cases.
None out of the box today. The platform ships one connector — Meta ads — and a lead endpoint any form can post to. A CRM connection is scoped and built as part of a deployment, and we would rather say that than print a logo wall.
The record is stored and you get an email. The assistant scores the lead against your criteria and writes the first reply, so the answer is waiting when a rep opens it.
A hard rule in the role: any price it was not given has to carry a pending-owner marker, and a draft that states a number without one is rejected before you ever see it.
It flags the lead instead of drafting a routine reply, and hands over the answers, the score and the reason behind it. Your rep picks up with the context already assembled.
No. Unistaff pricing is per AI employee per month. There are no per-message, per-minute, or per-lead surprise charges.