Vai al contenuto

Command Palette

Search for a command to run...

Back to Blog
product6 min read

Meet LIZZY: Our AI Prospecting Agent

Giovanni van Dam·

One Sentence to a Qualified Pipeline

Here's how most B2B sales teams find prospects today: open LinkedIn, search by job title and industry, scroll through results, visit company websites, check if they're a fit, look up decision-makers, find email addresses, log everything in a spreadsheet. Repeat for 2-3 hours.

Here's how it works with LIZZY: type "SaaS companies in the Netherlands with 20-100 employees that sell to healthcare" and get a qualified, enriched list of companies with decision-maker contacts. One sentence. One search. Done.

LIZZY is LeadScoutr's AI prospecting agent, and she's the reason we built the product.

How We Got Here

The idea for LIZZY came from watching sales reps work. Not from a whiteboard session or a competitor analysis — from sitting next to actual humans doing actual prospecting.

Every sales rep I observed during consulting engagements had their own research process. Some started with Google. Others started with LinkedIn. A few used industry directories. All of them spent between 60-90 minutes to research and qualify a single batch of 10-15 companies.

The process was always the same:

  1. Search for companies matching some criteria
  2. Visit each company's website
  3. Determine if they're actually a fit
  4. Find the right person to contact
  5. Get their email address
  6. Log everything somewhere

Every step was manual. Every step was repeatable. Every step was something AI could do better.

What LIZZY Does

LIZZY takes a natural language description of your ideal customer and runs the entire prospecting pipeline automatically.

Step 1: Understanding What You Want

You describe your target in plain language. Not filters, not Boolean operators, not dropdown menus — just words.

"B2B fintech startups in the US that have raised Series A funding and use Stripe."

LIZZY parses this into structured search parameters: industry (fintech), geography (US), business model (B2B), funding stage (Series A), tech stack (Stripe). She handles ambiguity well. If you say "mid-size companies," she interprets that as 50-500 employees. If you say "growing fast," she looks for recent funding and hiring activity.

Step 2: Searching the Web

Unlike traditional lead gen tools that query a static database, LIZZY searches the live web. She generates multiple search queries from your description, hits search APIs, and pulls back relevant company listings.

This matters because static databases have a shelf life. A company that launched three months ago isn't in last year's database. A company that just raised funding might not have updated their Crunchbase profile yet. But they probably have a website, a press release, and a LinkedIn presence. LIZZY finds them.

Step 3: Qualifying Each Company

For every company LIZZY finds, she visits their website and reads it. Not scrapes it for keywords — reads it. She extracts company descriptions, employee counts, product details, tech stack indicators, and recent news.

Then she scores each company against your ideal customer profile. The scoring model evaluates four dimensions:

  • Firmographic fit: Does their size, industry, and location match what you asked for?
  • Tech stack alignment: Are they using tools that indicate they'd benefit from your product?
  • Growth signals: Are they hiring, fundraising, or launching new products?
  • Engagement potential: Based on their current setup, how likely are they to need what you're selling?

Companies that don't meet the minimum threshold get filtered out. You only see qualified results.

Step 4: Finding the Right People

Once a company is qualified, LIZZY pulls decision-maker profiles. Names, job titles, verified email addresses, and LinkedIn URLs — the contacts you actually need to start a conversation.

She doesn't dump every employee at the company into your list. She identifies the people whose roles match your sales process. If you're selling to marketing teams, she finds the VP of Marketing and the Head of Growth. If you're selling to engineering, she finds the CTO and the Engineering Manager.

Why Conversational AI Beats Traditional Filters

Traditional lead gen interfaces are built for power users. Dropdown menus with 200 industry categories. Slider controls for employee count ranges. Multi-select boxes for technologies. It takes five minutes just to set up a search, and if you get one parameter wrong, you miss half your market.

LIZZY's conversational approach solves three problems:

Lower barrier to entry. Your CEO can use LIZZY. Your newest SDR can use LIZZY. Nobody needs training on how to set filters or what "NAICS code 511210" means.

Nuance that filters can't express. Try finding "companies that recently switched from HubSpot to Salesforce" with dropdown filters. You can't. LIZZY understands the intent and searches accordingly.

Iterative refinement. If the first results aren't quite right, you can refine in natural language: "Same search, but exclude companies with fewer than 10 employees." No need to rebuild the query from scratch.

Early Beta Results

We've been running LIZZY in private beta with a small group of B2B sales teams. Here's what we're seeing:

  • Average time to generate a qualified list: under 3 minutes (compared to 60-90 minutes of manual research)
  • Data accuracy: email verification rates consistently above 85%
  • User adoption: team members who previously avoided prospecting tools are using LIZZY daily — the natural language interface removed the friction

These numbers are early and will shift as we scale. We're sharing them because transparency matters more than polish.

What's Coming Next

LIZZY is the foundation. Here's what we're building on top of it:

Saved search templates. Define your ICP once, run it on a schedule, and get new prospects delivered automatically. Your pipeline refills itself.

Multi-turn conversations. Ask follow-up questions, drill into specific companies, and refine results through dialogue instead of starting over.

Lead scoring calibration. As you mark leads as won or lost, LIZZY's scoring model learns your specific patterns. The more you use it, the better it gets at finding companies like the ones you've closed.

Why We Built This

LIZZY exists because we watched too many good sales reps waste their best hours on research instead of conversations. The companies were out there. The data was out there. The process of connecting the two was just broken.

We didn't set out to build "the next AI tool." We set out to give small sales teams the prospecting power that used to require a dedicated research analyst and three different software subscriptions.

One sentence. One search. A qualified pipeline. That's LIZZY.

Condividi:

Resta aggiornato

Ricevi aggiornamenti sui prodotti e approfondimenti tecnici da Veldspark Labs.

Articoli correlati