You don’t need to code to build an AI agent. Start with the simplest, highest-value one — a follow-up email drafter that runs automatically after every meeting — using Lindy’s no-code builder. It takes about 30 minutes to set up, and you’ll see the value after your very next meeting.

Quick Verdict* Start with the email follow-up agent — it’s the lowest-risk, highest-value first build. * Lindy’s no-code builder gets you a working agent in about 30 minutes, no coding required. * Test thoroughly before going live — an agent that sends bad emails to clients is worse than no agent at all.

Try Lindy

What You’ll Build

| Part of the setup | What to do | Why it matters | | --- | --- | --- | | Trigger | Meeting ends (via calendar event completion or transcript availability) | Makes the agent run automatically without you lifting a finger | | AI Action | Draft a follow-up email summarizing the meeting, action items, and next steps | The core value — a ready-to-send draft in your Gmail within minutes | | Testing | Run the agent on a past meeting transcript before going live | Catches bad tone or missing details before a client ever sees them | | Next Agents | Virtual scheduler, meeting prep briefing, commitment tracker | Where to go once the first agent is running reliably |

How I Evaluated ThisI judged this setup on how fast a complete beginner can get from zero to a working, trustworthy agent — not on feature completeness. Every price below was re-verified in July 2026, and Lindy’s tier structure changed since this piece first published.

What Is an AI Agent (In Plain English)An AI agent is software that does things on its own, without you having to tell it every step. It’s different from a chatbot or a prompt in three ways:

1. It triggers automatically. You don’t have to open it and type something. It runs when something happens (a meeting ends, an email arrives, a calendar event fires).
2. It connects to your tools. It reads your calendar, checks your email, updates your CRM. It doesn’t just generate text in a chat window.
3. It takes action. It doesn’t just suggest what to do. It drafts the email, creates the task, books the meeting.

Think of it like this: ChatGPT is a smart friend you can ask questions. An AI agent is a smart assistant who does things for you without being asked.

Why Start With the Email Follow-Up AgentI’ve built dozens of agents. The email follow-up drafter is always where I tell beginners to start because:

• Immediate value. You’ll use it after your very next meeting.
• Low risk. It drafts emails. It doesn’t send them. You always review before hitting send.
• Simple to build. One trigger, one action, no complex logic.
• Impressive results. Prospects are blown away when they get a detailed follow-up two minutes after the call ends. They have no idea it’s automated.

What You’ll Need* A Lindy account — Lindy no longer has an ongoing free tier, but the 7-day trial is enough to build and test this agent * A Google account (for Gmail and Calendar) * A meeting tool that generates transcripts (Zoom, Google Meet, or a tool like Granola) * About 30 minutes

That’s it. Budget for Lindy’s Plus plan ($49.99/month) once your trial ends if the agent proves useful.

Step 1: Sign Up for Lindy (2 minutes)Go to lindy.ai and start the 7-day trial. That’s enough time to build and test this agent before deciding whether to subscribe.

Once you’re in, you’ll see the dashboard. Don’t get overwhelmed by all the options. We’re building one specific thing.

Step 2: Create a New Lindy (3 minutes)Click “Create New Lindy.” You’ll have the option to describe what you want in plain English or start from a template.

Type something like:

“After each meeting, read the transcript and draft a follow-up email summarizing what we discussed, any action items, and next steps. Save the draft in my Gmail.”

Lindy will interpret this and suggest a workflow. It usually gets about 80% right on the first try. You’ll fine-tune the rest.

Step 3: Set Up the Trigger (5 minutes)The trigger is what makes the agent run automatically. For this agent:

Trigger: Meeting ends (via calendar event completion or transcript availability)

Connect your Google Calendar. Lindy will ask for permission to read your calendar events. This is how it knows when a meeting happened and who was in it.

If you use a meeting notetaker (Granola, Otter, Fireflies), you can also trigger off the transcript being ready. Either approach works.

Pro tip: Set a filter so the agent only runs for external meetings. You probably don’t need follow-up emails after internal team standups.

Step 4: Configure the AI Action (10 minutes)This is the heart of the agent. You’re telling Lindy what to do when a meeting ends.

Action: Draft a follow-up email

Here’s the instruction prompt I use (customize for your style):

“Read the meeting transcript. Write a follow-up email from me to the other participants. Include:

– A brief thank you for the meeting

– 2-3 key points we discussed

– Any action items or commitments (who owes what)

– Suggested next steps

– Keep the tone friendly and professional. Keep it under 200 words. Don’t use jargon or overly formal language.”

Connect Gmail: Lindy will ask for permission to create drafts in your Gmail. It won’t send anything… just saves a draft for your review.

Step 5: Test It (5 minutes)Before trusting the agent with real meetings, test it.

Lindy usually has a “Test” button that lets you run the workflow with sample data or a past meeting transcript. Run it a couple times. Check the draft quality. Adjust the prompt if the tone or length isn’t right.

Common adjustments:

• “Make it shorter” if the drafts are too long
• “Use more casual language” if it sounds too formal
• “Always include a specific call-to-action at the end” if you want to drive next steps

Step 6: Turn It On (1 minute)Once you’re happy with the test output, flip the agent to active. From now on, after every qualifying meeting, a follow-up email draft will appear in your Gmail within a couple minutes.

Your job: open the draft, scan it, tweak a sentence or two if needed, hit send.

What Happens NextAfter a week of using your first agent, you’ll notice something. The 5-10 minutes you used to spend writing each follow-up? Gone. Across four meetings a day, that’s 20-40 minutes saved daily. But more importantly, you’ll actually send follow-ups consistently instead of procrastinating or forgetting.

That’s the real win. Not just speed. Consistency.

Common Mistakes Beginners MakeTrying to build something too complex first. Your first agent should do one simple thing. Not a 15-step workflow that touches six different apps. Start simple, then add complexity as you learn.

Not testing before going live. Always run test cases. An agent that sends weird follow-up emails to your clients is worse than no agent at all.

Writing vague instructions. “Write a good email” is vague. “Write a 150-word follow-up summarizing key discussion points and action items in a friendly tone” is specific. The better your instructions, the better the output.

Forgetting to review the output. AI agents draft things. You review them. Don’t blindly send AI-generated emails, especially early on. As you build trust in the output quality, you can spend less time reviewing.

Starting with a low-frequency use case. If something happens once a month, don’t automate it first. Start with tasks that happen daily or weekly. The time savings compound faster. That’s my 80/20 rule: automate the boring stuff that happens every day, not the impressive stuff that happens rarely.

Your Next Three AgentsOnce you’re comfortable with the email follow-up drafter, here’s the order I recommend building:

Agent 2: The Virtual Scheduler. CC the agent on scheduling emails. It checks your calendar, proposes times, and books meetings. I built a 3-minute delay into mine so people don’t realize it’s automated.

Agent 3: The Meeting Prep Briefing. Every morning at 7am, get a briefing with all your meetings, who you’re meeting, and context from past interactions.

Agent 4: The Commitment Tracker. If someone promises to send you something and doesn’t within two days, the agent drafts a polite reminder.

For the full breakdown of how these agents work together to save 15 hours a week, see my detailed agent setup guide.

Tools You’ll Need as You GrowLindy — Where your agents live. The 7-day trial gets you started. Plus at $49.99/month for serious use. This is what I use daily and what I teach in workshops.

Claude — When you want better writing quality in your agent outputs. Connect Claude as the AI model in Lindy for more nuanced drafts.

Zapier — Connects tools that Lindy doesn’t integrate with directly. Free tier gives you 100 tasks/month.

For a comparison of all the major agent platforms, check out my best AI agent platforms guide.

FAQDo I need to know how to code?No. Lindy is completely no-code. You describe what you want in plain English and connect apps visually. I’ve watched people in their 60s with no tech background build working agents in my workshops.

Is the 7-day trial enough to really test this?Yes. Lindy no longer has an ongoing free tier, but the trial window is enough to build the agent and run it across several real meetings — enough to know if it’s valuable before your first bill.

What if the email drafts aren’t good enough?Refine the prompt. Be more specific about tone, length, and what to include. You can also connect Claude as the AI model for better writing quality. Most people get good results after 2-3 prompt adjustments.

Can I build agents for my team?Yes, but start with yourself first. Build the agent, use it for a week, refine it, then share the workflow with your team. Lindy’s Plus plan supports multiple agents.

How is this different from Zapier automations?Zapier follows rigid if-then rules. AI agents can reason about what to do. A Zapier automation sends the same template email every time. An AI agent reads the actual meeting transcript and writes a unique, relevant follow-up. The intelligence is the difference.

Next StepTry Lindy