The inner kitchen of product creation is no less interesting and important than business development processes. That’s why we decided to introduce you to some exciting parts of development.
Rainex.GetLeads is a tool for finding and attracting new customers. In this article, we want to reveal to you the principles and key steps of lead prospecting and verification, so that you can make your own opinion about the value of automation in this area.
At GetLeads, we’ve developed an AI-powered system that helps businesses find and verify potential leads (ideal customers) via LinkedIn in real time. Our system combines web scraping, natural language processing, and machine learning to identify and validate leads based on custom criteria. Here’s how we built it.
The Starting Point
Rainex originally started out as a billing and subscription management system for customers. However, as we found out in our interactions with our clients, the main barrier to scaling was not recurring billing, but finding those who wanted to be billed.
We saw this as the perfect marriage of tools – find those who are ready to buy and get paid right away.
Since we were mainly focused on the B2B sphere, we took LinkedIn as a basis for our search. And these are the primary issues we paid attention to and wanted to solve:
- Manual searches are time-consuming
- Generic filters often return irrelevant profiles
- Job titles vary across companies and industries
- Verifying whether a person fits your ideal customer profile requires deep analysis
Our goal was to automate this process while maintaining high accuracy.
Our Solution: AI-Powered Lead Verification
Let’s now dive straight into the process of identifying a lead for prospecting and verifying it, which has 8 main stages. Here’s our step-by-step guide to the world of relevant potential customers:
Step 1: Generating the Ideal Lead Portrait
Users define their ideal lead using filters like:
- Location: Target countries and headquarters
- Company Size: Team size ranges
- Positions: Standard roles (e.g., CEO, CMO) or custom job titles
- Industry: Predefined sectors or user-defined customer descriptors
From these, our system generates a comprehensive lead portrait — a detailed profile of the ideal target.
Step 2: Extracting Key Phrases
Using NLP techniques, we analyze the lead portrait to extract:
- Key phrases and word combinations commonly found in relevant profiles
- Industry-specific terminology
- Synonyms and variations of job titles
This creates a rich set of terms for targeting and matching.
Step 3: Industry Classification
We map the lead portrait to one or more industry sectors.
For each sector:
- We maintain a curated list of job titles commonly found
- Users can customize these lists based on their use cases and experience
This classification allows for tailored lead matching.
Step 4: Scraping LinkedIn
With the extracted keywords and industry context, our system:
- Searches LinkedIn for relevant companies
- Extracts employee profiles from these companies
- Prepares a pool of potential leads
We use distributed scraping infrastructure to avoid rate limits and maintain speed.
Step 5–6: Profile Verification
For each profile:
- We classify the person into an industry sector
- Group profiles by sector to enable batch analysis
This ensures more consistent and context-aware evaluation.
Step 7: AI-Powered Position Verification
For each group of profiles within a sector:
- We generate prompts and send them to AI
- The AI determines whether the employee fits any of the target job titles for that sector
The model evaluates:
- Job title nuances (e.g., “Head of Growth” vs. “Growth Marketing Lead”)
- Work experience and descriptions
- Skills, endorsements, and profile summaries
The result is a fine-grained match, even for ambiguous or unique titles.
Step 8: Delivering Verified Leads
We present users with:
- A list of companies containing verified leads
- Specific employees who match the target profile
- Confidence scores to indicate match strength
This lets users focus only on high-confidence, validated prospects.
Technical Challenges and Solutions
When building Rainex.GetLeads, we set out to solve one of the most frustrating problems in B2B sales: wasting time on bad leads. Sales teams don’t need more contacts—they need the right ones, verified in real time, and ready to talk.
That sounds simple in theory. In reality? It took a lot of experimentation, failed ideas, and custom engineering. Here’s a look under the hood at the technical challenges we faced—and how we solved them using a combination of AI, smart systems, and stubbornness.
1. Title Variability: Building a Common Language for Job Roles
The challenge: Job titles in the wild are a mess. One company’s “Customer Success Rockstar” is another’s “Account Manager.” Multiply that by industry, region, and company size, and the data becomes chaotic. When you’re trying to identify decision-makers, this inconsistency breaks everything.
The solution: We built a custom title normalization system. It uses natural language processing to map unusual or creative titles to their standardized equivalents across industries and seniority levels. That means “Head of Growth” and “Marketing Strategist” can both be correctly classified as Marketing Decision Makers—without the system getting confused.
2. Scale: Scraping and Enrichment Without Getting Blocked
The challenge: To verify leads in real time, we needed to fetch and enrich thousands of data points per minute—without hitting rate limits or going dark due to blocked requests.
The solution: We developed a distributed scraping architecture that runs thousands of profiles in parallel. It uses rotating proxies, smart back-off timing, and region-specific routing to stay under the radar. The result is a scalable, reliable engine that continuously feeds fresh, verified data into our lead system with minimal downtime and rate limits.
The Impact
With these challenges tackled, GetLeads can now surface decision-maker leads that match your best customers, verified in real time, and scored for likelihood to convert. No manual research. No bad data. Just a pipeline filled with high-potential prospects.
It wasn’t easy—but if it were, everyone would be doing it. We just couldn’t wait around for “good enough” tools, so we built our own. And we’re pretty proud of it.
Our customers have achieved:
- Up to 90% reduction in lead search time
- 40% improvement in lead qualification accuracy compared to manual methods
- Higher lead discovery rates, especially in niche or underrepresented roles
What’s Next
While our current system already delivers strong results, we’re continuously improving it. Here’s what we’re working on next:
1. Faster Search & Results with a Pre-Built Database
Scraping LinkedIn in real time is powerful — but can be slow for large-scale searches. To improve speed and reduce latency, we’re building a proprietary database of pre-processed company and employee profiles.
This will enable:
- Instant lead retrieval for common ICPs
- Historical data tracking (e.g., role changes over time)
- Reduced scraping/API overhead
This hybrid approach will combine the flexibility of live search with the performance of pre-cached data.
2. Improving Accuracy with Feedback Loops
To continually refine our AI verification, we’re adding automated feedback systems and learning loops:
- User feedback integration — thumbs up/down on leads to inform model retraining
- Result re-ranking based on which profiles actually convert to meetings or sales
- Dynamic keyword tuning to reflect evolving job titles and industry terms
This will help the system self-optimize over time, especially for niche or fast-changing industries.
3. Outreach Analytics for Measurable ROI
Finding qualified leads is only half the equation — understanding what happens after outreach is critical for measuring success.
We’re adding a full analytics layer on top of our pipeline:
- Campaign performance tracking (open rates, response rates, click-throughs)
- Lead-to-customer conversion insights
- A/B testing for outreach messaging to see what resonates
These insights will give users visibility into what works — and help them continuously improve their approach.
Interested to experience how it works in practice?
Try out Rainex.GetLeads in a free trial and see first-hand how the tool will benefit and drive growth for your business.
Prefer well-organized guides? Let us walk you through the key points and help you immediately find the best settings for maximum results on a free demo.
Get Leads. Get new opportunities. Get a growth boost.
