Guides
A practical guide to building an AI recruitment workflow that screens CVs, ranks candidates, updates your tracker, and sends interview scheduling messages automatically.

Automating recruitment means using AI CV screening and automated interview scheduling to remove the repetitive admin between “new applicant” and “booked interview.” For founders, marketers, small-business owners, and ops teams, a good recruitment workflow automation can review resumes against a role scorecard, rank candidates, update your tracker, and send the next message without turning hiring into a black box.
The goal is not to let AI decide who gets hired. The goal is to give your team a faster, more consistent first pass so humans spend more time with qualified people and less time copying resume details into spreadsheets.
Recruiting has a lot of small tasks that feel harmless until they pile up:
Here is a realistic example.
A small company posts a marketing manager role and gets 120 applicants in one week. If an ops person spends just 3 minutes opening, saving, and logging each CV, that is 6 hours before any real evaluation happens. If the hiring manager then spends 8 minutes on each initial screen, that is another 16 hours.
With an AI hiring workflow, you might still review the top 30 to 40 candidates manually, but the admin and first-pass scoring can drop from 20-plus hours to 3 to 6 hours. At an internal cost of $40 per hour, that is roughly $560 to $680 saved on one role, plus faster replies to candidates. These are example numbers, not a guarantee, but they show where the payoff comes from.
The bigger win is consistency. Every CV is checked against the same criteria, every candidate gets a status, and your team sees the shortlist in one place.
This guide uses Impresiv as the visual, no-code canvas for the workflow. The automation will:
You can adapt the same structure for different roles, locations, departments, or seniority levels.

Start with the place where applications arrive.
Common trigger options in Impresiv:
For a simple small-business setup, use this:
Trigger node: Gmail, new email matching a label such as “Applications”
Input: Email body, sender email, attachments
Action: Save the CV attachment to Google Drive and pass the file into the next step
Tip: create one Gmail label per open role, such as “Applications: Sales Rep” or “Applications: Designer.” That makes routing much easier.
Next, the workflow needs readable text. Use the file from Gmail or Google Drive, then send the extracted text into an AI step.
Nodes involved:
Ask the AI to return only the fields you need. Keep it practical:
This creates clean data for the rest of the automation.
This is the most important part. Do not simply ask, “Is this candidate good?” That produces vague results.
Instead, define a role scorecard with clear criteria. For example, for a B2B content marketer:
| Criterion | Weight | Example signal |
|---|---|---|
| B2B writing experience | 30 percent | SaaS, agencies, professional services |
| SEO knowledge | 20 percent | Keyword research, briefs, rankings |
| Portfolio quality | 20 percent | Links, samples, measurable outcomes |
| Tool familiarity | 10 percent | CMS, analytics, automation tools |
| Role fit | 20 percent | Seniority, location, availability |
Then create an AI text model node that reads the CV and the job description, applies the scorecard, and returns structured notes.
Use a prompt like this:
You are assisting with first-pass CV screening for a hiring team.
Role: B2B Content Marketer
Job requirements:
- 3 plus years of B2B content or SaaS marketing experience
- Strong writing samples or portfolio links
- SEO experience, including briefs, keyword research, and content optimization
- Comfortable working with founders and small teams
- Bonus: experience with AI tools, automation, or technical products
Screening instructions:
- Evaluate only the information in the CV and application text.
- Do not infer protected characteristics or personal traits.
- Do not reject a candidate because information is missing. Mark it as missing.
- Score each criterion from 1 to 5.
- Give a total score out of 100.
- Recommend one of: shortlist, maybe, no for now.
- Provide 3 concise reasons for the recommendation.
- List 2 follow-up questions for an interview if shortlisted.
Return the answer in this format:
Candidate name:
Email:
Relevant experience summary:
Scores by criterion:
Total score:
Recommendation:
Reasons:
Keep the prompt role-specific. A sales role, operations role, customer support role, and engineering role should not use the same scorecard.
Once the AI generates its screening result, send it to a tracker.
Node: Google Sheets
Action: Add or update row
Suggested columns:
Use status values such as:
This prevents the classic hiring problem where half the team is working from email threads and the other half is working from a spreadsheet.
Now add conditional paths.
If recommendation is “shortlist”:
If recommendation is “maybe”:
If recommendation is “no for now”:
For sensitive hiring workflows, many teams choose to require human approval before any rejection email is sent. That is a smart default.
For shortlisted candidates, use Gmail to send a clear message.
Node: Gmail
Action: Send email
Message structure:
Example email:
Subject: Next step for the B2B Content Marketer role
Hi,
Thanks for applying for the B2B Content Marketer role. We reviewed your application and would like to schedule a first conversation.
Please choose a time using this scheduling link:
If none of the times work, reply with a few options and we will coordinate.
Best,
The hiring teamIf your team uses a scheduling platform, Impresiv can connect through HTTP/API or available integrations. If you prefer a lighter setup, send a booking link from Gmail and update the candidate’s status in Google Sheets when they reply.
You can also add a Slack notification:
“Shortlisted candidate invited: Name, role, score, CV link.”
Recruiting gets messy when people do not reply. Add a simple wait step and follow-up path.
Example:
This keeps the process moving without making your team chase every thread manually.
AI should summarize, score, and route. Humans should make hiring decisions. This is especially important for rejections, edge cases, and candidates with non-traditional backgrounds.
The better the scorecard, the better the output. Avoid vague criteria like “culture fit.” Use observable signals such as role experience, required certifications, writing samples, portfolio evidence, or tool experience.
Always keep a link to the original CV next to the AI summary. Hiring managers should be able to verify the recommendation quickly.
Before using the workflow live, run 20 past CVs through it and compare the results with your team’s actual decisions. Adjust the prompt, weights, and routing rules.
After an interview is booked, send a reminder through WhatsApp or Telegram with the time, meeting link, and what to prepare. This is useful for high-volume hiring or field roles where email response rates are lower.
Create a Slack workflow that posts a daily digest of new shortlisted candidates, including score, reasons, CV links, and suggested interview questions.
For creative, marketing, or technical roles, add a web scraping or HTTP/API step to inspect submitted portfolio links and summarize relevant projects for the hiring manager.
AI can make screening more consistent, but it still needs guardrails. Use job-related criteria, avoid protected or personal characteristics, log the AI’s reasoning, and keep a human reviewer in the decision loop.
Start with admin-heavy steps: saving CVs, extracting candidate details, updating Google Sheets, notifying Slack, and sending interview invitations. Add scoring once your role criteria are clear.
No. A simple stack like Gmail, Google Drive, Google Sheets, Slack, and Impresiv can handle a lightweight recruiting workflow. Larger teams can connect more systems through integrations or HTTP/API steps.
Usually not at first. A safer approach is to let AI draft or queue rejection emails, then require human approval before sending.
If you want to build this workflow without stitching together scripts, try it in Impresiv. Use the visual canvas to connect Gmail, Google Drive, Google Sheets, Slack, AI models, and scheduling tools through integrations or HTTP/API steps.
You can also use Magic Builder to describe the recruitment workflow you want and get a starting automation from a prompt. Sign up for Impresiv and build your first AI CV screening workflow today.