FlowsBook a callSign inCreate free accountSign up
AI employees and automation

An AI employee that reads job postings for you every morning

Instead of scrolling job boards, give one AI employee the roles, the area and what a good lead looks like. What it reads, what it keeps, and what reaches you each day.

Part of the guide AI employees

By the xMap team · · 5 min read

An AI employee that reads job postings for you searches public postings each morning for the roles you care about in your area, opens the ones it has not seen before, reads each description against your own note about what makes a good lead, and sends you only the companies worth a call, each with a score and a reason. You stop scrolling and start the day with a short list.

This article is for anyone who treats hiring as a signal: recruiters and staffing agencies, software and service companies selling to growing teams, training providers and consultancies. It explains what the employee does step by step, what you have to tell it, a made-up first week in Singapore, the setup mistakes that fill an inbox with noise, and how to decide how much it does alone.

What the employee actually does

In xMap Flows an AI employee is a flow with a job, shown as a role with a picture on your Flows page. One that reads postings is called a Hiring scout. Each run it works through the same steps a careful researcher would:

  • It searches. Job postings looks for live public postings matching your keywords in a city, or in the cities of the area you set, within a window such as the last 24 hours or the last 7 days.
  • It keeps one posting per company, by default, so a firm advertising ten roles appears once. Turn that off when agencies or large employers post many different roles you want to see separately.
  • It opens the details. With Get full details on, each posting brings the seniority, job type, industry, number of applicants and the full description.
  • It forgets what it has already sent. Only new lets through only companies this employee has not found before.
  • It reads. xMap LLM reads each description whole and writes a fit score from 0 to 100 and a reason, against your instructions.
  • It reports. It sends the short list where you work: an email, a Slack channel or a Google Sheets row per company.

What you have to tell it

The employee is only as good as its brief. Three things matter more than any setting:

  1. The roles, in the words employers use. "Finance controller" and "financial controller" may both be needed. Start with three or four phrases and look at what comes back before adding more.
  2. The area. A city, several cities, or a distance around a place. Postings are placed at the city named in them, so a posting listed only by a region sits at that city's centre on a map.
  3. What a good lead looks like, in plain words. For example: a company of 20 to 200 people, hiring its first person in this role, in our industries, not a recruitment agency. The AI scores each posting against that note, so a precise note gives sharper scores.

A made-up first week in Singapore

This example is invented. A small company in Singapore sells payroll software to growing firms. It hires a Hiring scout to look for companies advertising for HR managers, payroll officers and people operations leads, posted in the last 24 hours, with a note: firms under about 300 staff, hiring their first or second HR person, not agencies, not public bodies.

On Monday the scout finds 40 postings and keeps 26 companies after one posting per company. It scores them; nine reach 70 or more. The founder reads the reasons over coffee: one posting says the company is "building our HR function from scratch", another mentions moving off spreadsheets for payroll. Two high scores turn out to be agencies hiring on behalf of clients, so the founder adds a line to the note. From Tuesday the list is shorter and cleaner, and by Friday three of the companies have agreed to a demo.

Setup mistakes that fill the inbox with noise

  • Keywords that are too broad. "Manager" matches everything. Use the job titles that signal your need.
  • A window that is too long. Searching the last 30 days every morning brings old postings each time. Match the window to how often it runs, and let Only new do the rest.
  • No note about agencies. Recruitment firms post on behalf of others. Tell the AI how to treat them, or turn off one posting per company if they are your target.
  • Skipping the full details. Without the description, the AI can only judge the job title. Turn on Get full details when the score depends on the text.
  • Reading every result. Add a Filter on the fit score so only strong leads reach you.

How much it does alone, and what lands in your inbox

Every AI employee has a setting for how much it does on its own. Ask me first puts an approval before anything is sent outside your team. Act and tell me sends and then leaves a note. Fully on its own just works. For a scout that only reports to you, Act and tell me is usually right; if you later add Outreach to write to the companies, Ask me first lets you read each message before it goes.

The full shape, in order:

  1. Schedule: every weekday at 07:00.
  2. Job postings: your role phrases, your city, posted in the last 24 hours, with full details.
  3. Only new.
  4. xMap LLM: fit score and reason, against your note.
  5. Filter: fit score 70 or more.
  6. Email to you.

Each morning the email lists the strong leads: company, job title, a link to the posting and to the company's page, the city, the score and the reason. In the app the scout keeps a thread in the Team inbox with a daily note of what it found, and you can reply to it in plain words, for example "also look in Kuala Lumpur" or "pause until Monday". The templates Companies hiring near you and New companies hiring, daily are ready to hire as they are; Companies hiring buyers, and who to call adds the people to contact.

Questions

Can an AI read job descriptions and judge them as leads?

Yes. With full details on, xMap LLM reads each description whole and gives a score and a reason against the instructions you write, such as company size, industry or the words that signal a new team.

How do I stop the same company appearing every day?

Only new lets through only companies this employee has not found before, and one posting per company keeps a firm with many roles to a single line.

Will it pick up postings from recruitment agencies?

It can. Tell the AI in your note whether to score agencies down, or turn off one posting per company if agencies are the companies you want.

Where do the results go?

Wherever you choose: an email, a Slack channel or a row in Google Sheets. The employee also writes a daily note in its Team inbox thread.

Templates that do this

Read next

More on ai employees and automation · Every article · Templates

Put it to work

An AI employee in xMap Flows does this on live data, on its own timetable. Start from a template or describe it in plain words.