Automating Training Requests with a No-Code Intake Workflow

A training request intake workflow is a standard process, usually a form connected to automation, that captures every L&D request in one place, collects the information needed to assess it, and routes it to the right person. With today's no-code tools, a small team can build one in a day, and adding an AI step can turn a messy request into a structured summary before anyone reads it.

The real benefit isn't speed. It's that every request starts with the questions that matter: what problem are we solving, and how will we know it's solved?

Why intake matters

Most L&D teams receive requests like "Can we get a course on communication?" with no audience, deadline or goal. The first meeting is spent gathering basics. A good intake form does that work up front and quietly introduces stakeholders to performance-focused thinking, the same thinking behind action mapping.

The intake form: questions that matter

Keep it to about ten questions:

1. What business problem or goal is behind this request?
2. What should people be doing differently afterward?
3. Who is the audience, and roughly how many people?
4. Why aren't they doing it today? (Options: don't know how, process or tool issue, unclear expectations, not sure.)
5. How will we measure success?
6. Is there a deadline, and what's driving it?
7. Who is the subject matter expert?
8. Is there existing material?
9. Is this required by regulation or policy?
10. Who is the decision-maker?

Question four is the most valuable. It surfaces requests where training isn't the right fix.

The workflow, step by step

Here's the structure I use, which works with tools like Google Forms or Typeform for intake, Zapier, n8n or Power Automate for automation, and a sheet, Airtable or Notion database as the tracker:

1. Form submitted: the request lands in the tracker as a new row.
2. AI summary: an AI step reads the answers and produces a short summary, a suggested request type (new course, update, job aid, not a training issue) and a list of missing or unclear information.
3. Acknowledgment: the requester gets an automatic email confirming receipt, the expected response time and any follow-up questions the AI flagged.
4. Team notification: a message in Slack or Teams with the summary and a link to the record.
5. Human triage: an L&D lead reviews, prioritizes and assigns. The AI suggestion is input, not a decision.
6. Status updates: when the status changes in the tracker, the requester is notified automatically.

This is the same pattern I used in the Smart Onboarding Pipeline project: one form, AI-generated briefs and automatic notifications, replacing a long manual checklist.

A sample AI summary prompt

You are an L&D intake assistant. Read this training request. Output: (1) a two-sentence summary, (2) the likely request type: new training, update existing, job aid, or likely non-training issue, with a one-line reason, (3) up to three clarifying questions for missing or vague answers. Do not make approval decisions. [Form responses]

Keep the AI step on an approved tool, and avoid sending sensitive personal data through it. My notes on safe AI use at work apply here too.

Mistakes to avoid

• Too many questions. If the form takes 20 minutes, people will go back to emailing you.
• Letting AI decide priority. Priority involves politics, strategy and capacity that a model can't see.
• No feedback loop. If requesters never hear back, they'll stop using the form.
• Skipping the "not training" path. Have a polite, useful response ready, such as a job aid or a process conversation, for requests that don't need a course.

Where to start

Build the form first and use it manually for two weeks. Once the questions feel right, add the acknowledgment email, then the notification, then the AI summary. Adding automation one step at a time keeps it easy to fix.

Sources

• Cathy Moore, Action mapping: A visual approach to training design
• NIST, AI Risk Management Framework