tariqdeir.com / writing

· ai · 5 min

I built an AI agent that qualifies my discovery calls automatically

How I stopped checking my inbox and started showing up to calls already knowing who I'm talking to—and whether they're a fit.

TL;DR

  • Discovery calls are time-intensive, most founders waste hours on unqualified prospects because they don't know who they're talking to until the call starts.
  • I built a Lindy agent that automates qualification, it scrapes LinkedIn profiles, analyses fit against my ICP, and delivers a custom prospect report to Slack before every call.
  • The system saves me 3–5 hours per week and ensures I show up to every call prepared, focused, and already knowing the conversation openers that matter.

Here's how discovery calls used to work for me:

Someone books a slot on Calendly. I get a calendar invite. Maybe I glance at their email address before the call. Maybe I don't. Either way, I'm flying blind until we're five minutes into the conversation.

That approach is fine when you've got time to burn. But when you're building One Task Today alongside a full-time job, a family, and multiple ventures, every minute counts.

I needed a system that gave me the information that matters, before I ever join the call.

The problem: discovery calls are a black box

Discovery calls are supposed to be mutual exploration. But too often, they're one-sided interrogations where I'm scrambling to understand who this person is, what they do, and whether we're even a fit.

The typical flow looked like this:

  1. Prospect books a call
  2. I receive a calendar notification (buried in my inbox)
  3. I join the call with zero context
  4. First 10 minutes = "So tell me about yourself..."
  5. Realise halfway through they're not remotely a fit

Rinse. Repeat. Hours wasted.

All the information I needed was already available. LinkedIn profiles. Company websites. Recent posts and wins. I just wasn't capturing it systematically.

The solution: an AI agent that does the work for me

I built a simple but powerful agent on Lindy that handles three critical tasks automatically:

1. Meeting details → Slack notification

As soon as someone books a discovery call, the agent sends the meeting details to a dedicated Slack channel. No more checking my inbox for calendar invites. Just one clean notification in the channel I actually monitor.

2. LinkedIn profile scraping

The agent scrapes the prospect's LinkedIn profile and pulls the key data points: current role, company, recent activity, mutual connections, and anything else that gives me context on who they are and what they're building.

3. Custom prospect report

Here's where it gets good. The agent takes all that scraped data and runs it through a qualification framework. It has complete context on:

  • My business
  • My goals
  • My ICP
  • My qualification criteria

Then it generates a custom report that includes:

  • Who they are: Role, company, stage of business
  • What they do: Core offering, target market, business model
  • How they do it: Team size, funding status, go-to-market approach
  • Recent wins: Launches, milestones, notable achievements
  • Conversation openers: Specific talking points based on their activity and challenges
  • Fit assessment: Clear recommendation on whether they match my ICP

All of this lands in Slack before I ever join the call.

Why this changes everything

The difference between showing up to a call blind and showing up prepared is massive.

Before the agent:

  • First 10 minutes wasted on basic discovery
  • No context on fit until deep into the conversation
  • Generic questions that don't land
  • Feeling reactive instead of in control

After the agent:

  • I know exactly who I'm talking to before I join
  • Conversation openers are specific and relevant
  • I can disqualify bad fits before the call even happens
  • High-quality prospects get a better experience because I'm prepared

The agent doesn't just save time. It makes me better at my job.

What this really means: focus on what matters

The beauty of this system isn't just automation for automation's sake. It's selective attention.

I don't need to check my inbox obsessively anymore. I don't need to context-switch between email, calendar, and LinkedIn. I get one notification, in one place, with all the information I need to decide whether this call is worth my time.

And if it is? I show up ready to deliver value immediately.

This is the edge that matters in 2025: knowing what deserves your attention and filtering out what doesn't.

Building this yourself

You don't need to be a developer to build something like this. Lindy is a no-code AI agent platform that lets you chain actions together with simple logic.

The basic flow looks like this:

  • Trigger: New calendar event created (via Calendly or Google Calendar integration)
  • Action 1: Extract attendee email and meeting details
  • Action 2: Scrape LinkedIn profile using the email (or name + company)
  • Action 3: Run the scraped data through an AI prompt that includes your ICP, business context, and qualification criteria
  • Action 4: Send the formatted report to Slack

Total setup time? Less than an hour if you know what you want.

The hard part isn't the tool. It's clarity on your qualification criteria. If you don't know what makes someone a good fit, the agent can't help you.

The broader pattern: agents as operating leverage

This isn't just about discovery calls. It's about a shift in how we work.

We're moving from "doing everything manually" to "orchestrating agents that do the repeatable work for us." The goal isn't to eliminate human judgment—it's to preserve human judgment for the decisions that actually matter.

I don't need to manually qualify every prospect. I need to show up prepared for the ones who are a fit and gracefully decline the ones who aren't.

That's leverage.

And the best part? This is just the beginning. Every founder should be asking: What am I doing repeatedly that an agent could handle better?

Action steps

  1. Audit your discovery call process right now. How much time are you spending on unqualified prospects? How often do you show up to calls unprepared?
  2. Define your ICP and qualification criteria clearly. Write down the specific characteristics that make someone a good fit vs. a waste of time.
  3. Identify one repeatable task that's eating your time. Discovery call prep, email triage, client onboarding—pick one and explore how an AI agent could automate it.
  4. Set up a Lindy account and experiment. Start with a simple automation (like calendar → Slack notifications) and build from there.
  5. Review after two weeks. How much time did you save? What would you automate next?

The bottom line

Discovery calls don't have to be a black box. With the right automation, you can show up to every conversation already knowing who you're talking to, whether they're a fit, and exactly what to ask.

The question isn't whether AI agents can help you. It's whether you'll build the systems that let them.