Skip to content
AI Lead Factory · running live

Your own AI sales team. It finds buyers, writes to them and books the calls.

An autonomous outbound system that researches your market, picks the offer buyers want now, finds decision-makers, writes one personal email at a time and reads every reply. You deliver the work; it fills the calendar.

If you received an email from Doutomation, this system researched you, wrote it and sent it.

1
person per company
≤ 5 min
to honour an opt-out
Mon–Fri
business-hours sending
100%
yours: domain, data, tools
Who built it

Engineered by hand, by someone who does this for a living.

I'm Do Anh, a Certified Zapier Expert and Top Rated Plus automation engineer with a degree in Software Engineering. Since 2019 I've built automations for businesses across the US and Canada. The Lead Factory is the system I designed to find my own clients, and it runs on my pipeline every day.

“You could build something like this yourself. You would also hit every bottleneck I hit: burned domains, silent data gaps, duplicate sends, fake ‘leads’ from auto-replies, runaway AI costs. I've already found and fixed them.”

Not a template, and not a one-click AI generator. Every agent, rule and limit was designed, tested and tuned by hand, and every safeguard exists because that failure happened in real life.

630
automation jobs delivered
$300K+
earned on Upwork
280
client reviews
2019
building automations since
Bottlenecks you'd hit building it yourself
  • Your domain gets flagged as spam after a week
    Dedicated warmed-up domain, low daily caps, one person per company, random send times inside business hours.
  • The data provider quietly returns nothing
    Empty pages are retried, never cached, and filters are checked so the right industry is really searched.
  • A crash sends the same email twice
    Every email gets its ID before it is sent and moves through one-way states, so a retry can never resend it.
  • Out-of-office replies counted as leads
    Header checks plus AI intent separate real people from auto-replies and bounces.
  • Opt-outs slip through
    Replies are read within minutes and opt-outs are suppressed permanently, across every campaign.
  • AI costs creep up unnoticed
    Every AI call and data credit is reserved against a hard monthly cap and settled with the real cost.
  • Long AI calls time out and get billed twice
    Timeouts tuned for deep reasoning, no hidden retries, and a retried step replays its saved plan.
  • Emails land at 7 PM for the buyer
    Sending follows the buyers' time zone, weekdays only, spread at random through the working day.
Why outbound breaks

Good businesses stay small because nobody has time to sell.

Hiring an SDR

Months to hire and ramp, one person's hours, and the pipeline stops when they leave.

Lead-gen agencies

Big lists, generic templates, your domain on the line, and the leads stay theirs.

Doing it yourself

Another dashboard, another tool, and outreach is the first thing to slip when work gets busy.

Automation in motion

Don't read about it. Run it.

The actual pipeline, step by step. Press Play or step through it: the run log shows what each station does and the inspector shows the data it passes on. Steps are real; the data is made up for the demo.

Ready
Run log
  1. Press Play to run the workflow.
Data inspectorexample values
{ }
Meet your AI sales team

Seven agents. One job each.

Behind the flow above is a small team of AI agents that works like a sales department: a director who plans, specialists who research, prospect, write and read replies, and a gatekeeper that enforces the rules. They hand work to each other through a shared queue and a shared memory, around the clock.

The Director

Head of sales

Reads the whole pipeline every few hours: budget, campaigns, leads, replies, bookings and the latest market brief. Then it writes a plan.

Decides
Which offer to sell to which segment, when to launch, test, pause or retire a campaign, and where to shift volume.
Never
Send an email itself, raise a limit or skip a safety gate.

Market Analyst

Research

Searches the live web once a week and writes a sourced brief: what buyers are paying for, which industries feel the pain, typical prices, who signs.

Decides
Which experiments look hottest right now, ranked for the Director.
Never
Invent a number or follow instructions hidden in a web page.

Offer Strategist

Positioning

Turns what you actually deliver into concrete offers: the pain, the outcome, the buyer titles, the company profile.

Decides
How your service is framed for each segment.
Never
Promise something you have not delivered before.

Lead Hunter

Prospecting

Plans its own search strategies on B2B data providers, keeps what yields verified decision-makers and drops what does not.

Decides
Which titles, industries and company sizes to search, and when to stop spending credits.
Never
Keep an unverified email or contact two people at one company.

Copywriter

Outreach

Writes one short email per person from their role, company and your offer, signed by you.

Decides
The angle and wording for each prospect, and whether the offer fits them at all (it skips weak fits).
Never
Send without passing the gates, or drop the opt-out line.

The Gatekeeper

Compliance, plain code on purpose

Checks every email before it leaves: suppression list, one person per company, daily cap, business hours, budget.

Decides
Pass or hold. Deterministic rules with no AI judgement, so nothing can talk its way past them.
Never
Make an exception.

Inbox Analyst

Replies

Reads every reply within minutes, tells real people from auto-replies and bounces, and sorts intent: interested, question, not now, referral, no.

Decides
What each reply means and what happens next: stop the sequence, suppress, or flag a hot lead to you.
Never
Count an out-of-office as interest or ignore an opt-out.

And you

You only see what needs a human: interested replies and booked calls. Tell the Director what you want in plain English, like "find 20 more leads in insurance" or "pause the B test", and it plans the steps.

Hire the team
Why it works

Built like a sales team, governed like software.

Most AI outreach tools are a template and a mail merge with a language model on top. This one is designed the other way round: judgement where judgement helps, hard rules where mistakes are costly.

  1. 01

    A team of specialists, not one chatbot

    Each agent has one job and a narrow set of tools. A copywriter that only writes and a researcher that only researches make fewer, smaller mistakes than one model asked to do everything.

  2. 02

    AI decides, code enforces

    Strategy, targeting and wording come from AI. Sending limits, suppression, compliance and budget are hard rules in code that no agent can override.

  3. 03

    It remembers and measures

    Every decision, email, reply and cent is recorded. Each cycle starts from what actually happened, not from what a prompt assumes.

  4. 04

    Driven by the market, not a template

    Offers follow fresh research on what buyers want this month, and the Director keeps testing two at a time until one clearly wins.

  5. 05

    A human stays in charge

    I set the objective and limits, watch the agents work in a live control room and can redirect them in plain English. They do the repetitive work; judgement stays human.

What makes it different

Not a mail merge. A sales system.

Ten things it does that a template tool, a list broker or a busy founder can't keep up every day.

It thinks before it sends

An AI Director chooses which offer to sell to which segment, runs two side by side and retires what stalls. Every decision is logged with its reasoning.

It reads the market every week

Fresh web research on what buyers are paying for now, so your outreach follows real demand instead of last year's playbook.

One person, one email

Each email is written for one decision-maker, and only one person per company is contacted. No blasts, no lists of 10,000.

Paced like a human

Sends at random times inside your buyers' business hours, under a daily cap per mailbox, from your own domain.

Reads every reply

Auto-replies are never counted as interest. Real replies are classified by intent, and opt-outs are honoured within minutes.

Compliance built in

Suppression list, opt-out line, physical address, verified emails only. A message that fails a check is held, not sent.

A hard budget cap

Every AI call and data credit is metered against a monthly cap. You see cost per lead and, as replies come in, cost per booking.

You own everything

Your domain, mailboxes, data provider account and every lead. If you leave, the pipeline and its history stay yours.

Supervised, not left alone

Every agent works in a live control room I watch, and every decision is logged with its reasoning, so nothing happens that can't be explained.

Steer it in plain English

"Find 20 more leads in insurance." "Pause the B test." Instructions go through the same safety gates as everything else.

Compared

How it stacks up

CriterionDo it yourselfHire an SDRLead-gen agencyAI Lead Factory
Researches demand before choosing an offer
Personal email per decision-maker
Reads and sorts every reply
Works every weekday without ramp-up
Protects your domain (caps, one per company)
You own data, domain and tools
Transparent cost per step
Typical monthly costYour timeSalary + toolsRetainer + lock-inFrom $1,500
Pricing

Start with a pilot. Keep what works.

No long contracts. Tool costs (domain, mailboxes, data provider) are usually under $100 a month and billed to you directly, so you own the accounts.

Pilot

Start here
$1,500
one-time · 30 days
  • Domain, mailbox and DNS setup (SPF, DKIM, DMARC) plus warm-up
  • Your service turned into offers and ideal-customer profiles
  • One offer × one segment, about 300–400 personal emails
  • AI reply handling and opt-out management
  • End-of-pilot report: sends, replies, bookings, cost

Credited to your first month if you continue.

Start a pilot

Growth

From $1,500
per month
  • Two campaigns side by side (A/B)
  • About 1,000 personal emails a month
  • Calls booked straight onto your calendar
  • Weekly market brief and performance report

After the pilot. Cancel any month.

Talk about Growth

Scale

From $3,000
per month
  • Up to four campaigns and several sending domains
  • About 3,000 personal emails a month
  • Director rotates offers and segments on fresh research
  • Daily report and a monthly strategy review

For several offers, regions or brands.

Talk about Scale
The pilot

Thirty days, four steps

  1. Week 1

    Set up and warm up

    We register or connect your sending domain, set up mailboxes and DNS, and start warm-up. You share what you sell and who buys it.

  2. Week 2

    Offers and first sends

    The system turns your service into offers and buyer profiles, finds verified decision-makers and starts sending at a careful daily pace.

  3. Weeks 3–4

    Replies, calls, learning

    Replies are read and sorted, interested buyers book calls, and the Director shifts volume towards what works.

  4. Day 30

    Report and decide

    You get the real numbers: emails sent, human replies, bookings and cost. Continue on Growth or Scale, or keep everything and stop.

Fit

Built for businesses that sell to businesses

A great fit

  • Wholesalers and manufacturers selling to retailers, hotels or distributors
  • B2B service firms: accounting, IT, logistics, marketing, staffing
  • Agencies and consultancies with a clear, specific offer
  • Software and SaaS companies selling to small and mid-sized businesses
  • You can take a few more good calls a week

Not the right tool

  • Reaching consumers at home (B2C cold email)
  • Businesses that need thousands of emails a day
  • Offers that can't be explained in two sentences yet
  • Regulated outreach that needs prior consent
Questions

Straight answers

Is this spam?

No. Every email is written for one named decision-maker at a business, only one person per company is contacted, sends are capped and spread through business hours, every email carries an opt-out line and your address, and opt-outs are suppressed permanently. It follows US CAN-SPAM rules for B2B email.

Who is it for?

Businesses that sell a specific product or service to other businesses: wholesalers and manufacturers, B2B service firms, agencies, software companies. The more concrete your offer, the better the emails. It is not built for reaching consumers at home.

What results can I expect?

We don't publish made-up reply rates. The pilot exists so you see your own numbers within 30 days: emails sent, genuine human replies, bookings and cost per booking. Live figures from our own pipeline will appear on this page as they accumulate.

Will it burn my domain?

We send from a dedicated domain that we warm up first, keep volume low per mailbox, verify every address, and stop on bounces and opt-outs automatically. Your main business domain is never used for cold email.

What do I need to provide?

A clear description of what you sell and who buys it, a calendar link, and accounts in your name for the domain, mailboxes and the B2B data provider (we help set them up). Tool costs are typically under $100 a month and paid directly by you.

Is a person involved?

Yes. I set up and supervise your pipeline, review the Director's decisions and step in on anything unusual. The AI does the repetitive work; the judgement calls stay human.

Let the factory sell. You do the work you love.

A 30-minute call to see if your offer fits, what the first campaign would look like and what it would cost. No pitch deck.