Back to Cases
#arrow-left

Unibrix

Engineering / Knowledge Base

#menu

Cases / Automation / Workflows

BLOG_ARCHIVE

#logo
#close
Start building now
#arrow-right
Get in touch
#envelope
LET'S BUILD

Ready to start your project?

Tell us about your challenge. We'll engineer a solution that scales with precision.

#envelope
EMAIL
contact@unibrix.io
#calendar
SCHEDULE
Book a 30-min consultation
We typically respond within 24 hours
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Close
PROJECT OVERVIEW

ICP Filtering

[ SALES AUTOMATION & LEAD INTELLIGENCE ]

End-to-end automated pipeline for qualifying B2B leads by industry — built on n8n, powered by AI classification, and operated entirely from Slack.

CATEGORY
Sales Automation / Lead Intelligence / B2B
TIMELINE
Active
COMPONENTS
n8n workflows, Slack bot, Google Sheets, AI classification layer
STATUS
● DELIVERED
SERVICES PROVIDED
Workflow Automation
AI Integration
Slack Bot Development
Data Pipeline Engineering
Process Design
Execution Monitoring
TECHNOLOGY STACK
n8n
Slack
Google Sheets
Google Drive
AI / LLM
HTTP APIs
DataTable
OAuth2
n8n workflow for AI-powered lead qualification, website classification, and automated B2B sales prospecting
01
STEP 01 — Project overview

Assembly Requirements

MODULE INVENTORY (5 COMPONENTS)

Qualifying leads at scale is one of the most time-consuming bottlenecks in B2B sales. Reviewing hundreds of company websites manually — determining whether each one is a fintech, marketplace, edtech, or healthtech — is slow, inconsistent, and impossible to do at volume without growing the team.

Unibrix built the ICP Filtering System: a fully automated pipeline that takes a raw Google Sheet of companies, classifies each one by industry using rule-based logic and AI, and delivers a clean results spreadsheet — all triggered and monitored from a Slack bot, with no engineering involvement required.

Primary focus
#dot
Automating manual lead qualification at scale
#dot
Multi-industry classification with high accuracy
#dot
Self-service tooling for non-technical sales teams
#dot
Real-time monitoring and execution control
What we delivered
#dot
End-to-end ICP filtering pipeline on n8n
#dot
AI-powered fallback classification layer
#dot
Slack bot control center with live monitoring
#dot
Product catalog and browsing system
#dot
Execution tracking with stage-level reporting
#dot
Support for fintech, healthtech, edtech, marketplace
02
STEP 02 — CHALLENGES IDENTIFIED

Assembly Requirements

MODULE INVENTORY (5 COMPONENTS)
REQUIRED CAPABILITIES
A
#checked
Scale
Manual review of company websites doesn't scale. Sales teams were spending hours filtering leads that should take minutes — and accuracy degraded under volume pressure.
B
#checked
Accuracy
Rule-based classification alone couldn't handle non-English sites, ambiguous business models, or the nuance between a marketplace platform and an online store. An AI fallback layer was required for edge cases.
C
#checked
Visibility
Sales teams had no real-time insight into filtering progress, execution status, or results without involving engineering. The system needed to be fully self-service with live feedback.
03
STEP 03 — ASSEMBLY PROCESS

How We Built It

MODULE INVENTORY (3 COMPONENTS)
MODULE 3.1
● PIPELINE ARCHITECTURE

A three-layer architecture separates concerns cleanly: a Slack Bot as the user interface, a Main Workflow as the orchestrator handling data ingestion, batching, and deduplication, and an Executor Workflow responsible solely for per-website classification logic.

Pipeline architecture for n8n sales automation, AI lead qualification, and automated website validation workflow
MODULE 3.2
● CLASSIFICATION ENGINE

The executor performs industry-specific checks per domain: TLD shortcuts, HTTP validation, keyword matching, platform detection, and business model scoring. An AI classification layer activates as a fallback for non-English sites — improving accuracy on ambiguous international leads.

AI-powered classification engine in n8n for website analysis, industry detection, and automated B2B lead qualification
MODULE 3.3
● SLACK CONTROL CENTER

A full-featured Slack bot gives sales teams complete control without touching n8n. Users can trigger runs, monitor live executions, inspect results, stop active workflows, and pull historical statistics — all via slash commands and interactive buttons. Duplicate-run protection surfaces a confirmation modal before starting parallel executions.

Slack dashboard for managing n8n workflow automation, AI lead qualification, and sales automation executions
04
STEP 04 — ASSEMBLY COMPLETE

Delivered Impact

MODULE INVENTORY (5 COMPONENTS)
~2 Minutes
PROCESSING SPEED
100 companies fully classified
4 Verticals
INDUSTRIES COVERED
Fintech, Healthtech, Edtech, Marketplace
Self-Service
TEAM ACCESS
Zero engineering involvement
PERFORMANCE OUTCOMES
A
#checked
Eliminated manual lead qualification from the sales workflow entirely
A
#checked
Enabled non-technical sales teams to run and monitor ICP filtering independently
A
#checked
Delivered stage-level transparency into every classification decision
A
#checked
Established a scalable foundation ready for additional industry verticals
ASSEMBLY PHILOSOPHY
We don't just code.
We build systems.
Piece by piece,
designed to snap perfectly
into your vision.
#dot-circle
#dot-circle
#dot-circle
#dot-circle
Start building now
#arrow-right