1. Home
  2. Work
  3. StudyfinderAI outreach

Outreach on fixed rules

Two founders and more than 6.000 schools and universities on the list. From March to November 2026 the system found the right contact, sent, followed up and sorted the replies. AI proposed. Fixed rules decided. A person took every warm reply. It's paused for now.

Business
StudyfinderAI, co-founded
My role
Designed and built it
Ran
March to November 2026
Status
Paused
4.199

emails sent over 85 send days

What it is: first emails and follow-ups to schools and universities, March to November 2026.

Why it matters

every one came from a template a person signed off.

Step 06, Send and follow up
1.163

institutions contacted

What it is: schools and universities that got at least one email.

Why it matters

the system found the right contact on each institution's own website first.

Step 02, Contact research
19

countries

What it is: the countries the institutions are in.

Why it matters

each country has its own holiday calendar. While a school is closed, nothing goes out there.

Step 03, Holiday check
100

emails a weekday, hard limit

What it is: the most the system may send in one day.

Why it matters

a hard limit protects the sending domain. Sending also pauses by itself when bounces pass 5%.

Step 05, Send gates

Counted from the system's own database, 23 March to 9 November 2026. I don't show reply rates. The count isn't clean enough to publish.

How it ran

Eight steps, every weekday it ran

  • AI
  • Fixed rules
  • A person decides

The crew

Three agents, one job each

I called the system Project Nebuchadnezzar and named the agents after the crew of that ship in The Matrix. Each one has a written boundary: what it owns, and what it must never touch.

  • Morpheus, the orchestrator. Reads the state of everything, decides the next job, starts it and keeps watching it. It talks to me in one message thread on my phone: a digest at the end of the day, and exceptions only.
  • Link, the researcher. Imports institution lists, fills gaps such as website and city, finds and scores contacts, and prepares the holiday calendar per country. It never sends an email.
  • Neo, the mailer. The only agent that touches the mailbox. It checks the gates, sends, plans the follow-ups and sorts the replies.
  • A control room to watch them. A small game-style world where the agents walk around, a notice board shows the open tickets and a dashboard shows backlog, blockers and how fresh the data is. It only reads: nothing in it can send or change data.

351 commits from March to November 2026. The language model runs in the cloud first, with a local model on the always-on computer as backup. It never decides whether to send.

Step by step

What happens at each step

01Batch import

A person decides

A list of institutions comes in. The system previews it on a copy of the database first. Nothing is saved until I approve the preview.

02Contact research

AI advises, rules decide

  • The system looks for a contact on each institution's own web domain and scores it.
  • It saves a contact by itself only above a fixed score: 85, or 80 for a role address such as admissions.
  • AI writes a short line on why this contact fits. That line is advice for me, never the decision.
  • In one run, 60 contacts were filled in automatically from 127 candidates. The rest waited for a person.
03Holiday check

Fixed rules

A table of holiday periods per country, taken from official calendars. While a closure runs, nothing goes out to that country. An email to an empty office is a wasted first impression.

04Orchestrator

Fixed rules

Every 15 minutes in working hours one job checks the state of everything and starts the next task: research, sending or inbox. No model decides the order.

05Send gates

Fixed rules

An email goes out only if every gate is open:

  • a valid contact exists and nothing is flagged for review;
  • no holiday is running in that country;
  • it's inside the send window: Monday to Friday, 08:00 to 12:00 and 13:00 to 14:00;
  • the limits hold: 25 per run, 25 per hour, 100 per day, 30 seconds between sends;
  • the spam check passes.
06Send and follow up

Fixed rules

  • A first email, then up to 4 follow-ups after 14, 7, 7 and 7 days.
  • Written templates in English and German, with and without a name, and an opt-out line in every email.
  • AI drafting exists in the system and is switched off. Every email comes from a written template that a person signed off.
07Inbox triage

AI labels, rules act

  • Out of office: the follow-up moves to after the return date.
  • Bounce: the contact is marked and the research step looks for a replacement.
  • Removal request: the institution goes on a do-not-contact list, for good.
  • Anything unclear: AI adds a label, and the reply waits for a person.
08Handoff and reporting

A person decides

A clear yes gets a short, warm answer from a fixed template, with me in copy. From there I take over. A mixed or unclear reply comes to me directly. Once a week I get a report with the numbers. Alerts only come for exceptions, so I only look when something needs a decision.

Guardrails

What keeps it safe

  • AI never presses send. It researches, labels and summarises. Fixed rules decide what goes out.
  • Automatic brake. Sending pauses by itself when bounces pass 5%, to protect the sending domain.
  • Test and live are separate. Two databases. The system refuses to run when they don't match.
  • Dry runs and backups. A dry-run mode, and a database backup before every automatic contact update.
  • New subject lines need my yes. AI may suggest them. I approve them.

Built with

Stack, in plain words

  • Three Python agents: research, mailing and an orchestrator.
  • A small SQLite database, the Gmail API and a contact finder service.
  • A language model for advice and labels, with a local model as fallback.
  • An always-on computer with scheduled jobs and phone alerts.
  • About 880 automated tests that run on every change.

What went wrong

What broke, and what I changed

  • Auto-replies counted as real replies. Out-of-office mails and bounces were logged as replies, and some were even handed on as warm leads. I rebuilt the sorting to spot them in any language. It's also why I don't publish reply rates.
  • School holidays shut out universities. The holiday check used school calendars for universities too, so German and Austrian universities got nothing for three weeks. I reopened them. A closure has to know what kind of institution it applies to.
  • Old templates went out. 37 cold emails to schools used the old text. I set that lane to zero sends a day until the new copy was approved.

For your business

What you can take from it

  1. Let AI propose, let rules decide.

    Use AI where judgement is cheap to check: research, labels, summaries. Keep the send button, the price and the promise in fixed rules.

  2. Build the brake before the engine.

    Hard limits and automatic pauses protect what is hard to win back, like a sending domain or a customer list.

  3. People see only the exceptions.

    A weekly report and alerts for the unusual. Your team spends its time on the replies that matter.

Sources: the system's configuration, decision log and database, read 2026-11-16. The numbers describe my engineering work, not StudyfinderAI product results.

Contact

Tell me what is stuck

Book 30 minutes on video, or write me a few lines about which AI tools your team uses and where it gets stuck.

Book a 30-minute call

consulting@mario-behrends.com
LinkedInEnglish and German.