Alpha Lead Engine
From prospect discovery to closed conversation — fully automated, human-reviewed.
Overview & Impact
The Problem & The System Solution
The Problem
Outbound sales teams spend enormous time discovering companies, checking lead quality, researching websites, writing personalized emails, managing follow-ups, and manually reading replies.
The System Solution
Designed a unified AI-powered pipeline that moves each prospect from discovery to qualification, personalization, outreach, follow-up, reply classification, and campaign analytics — with a mandatory human review checkpoint before any email is sent.
Sales development is repetitive and expensive. SDRs spend 60-70% of their time on tasks that don't require human judgment: finding companies, checking if they fit the ICP, researching their website, and writing a first email that sounds personal.
Alpha Lead Engine automates exactly those tasks while keeping humans in control of what actually matters: reviewing and approving outreach before it reaches a real prospect.
The system is modular by design. Each of the 6 AI modules has a single responsibility. This makes it testable, replaceable, and improvable independently — a direct contrast to monolithic AI automation systems that fail unpredictably.
Architecture & System Flow
System Design Visualizer
Alpha Lead Engine 6-Module Flow
Step-by-Step System Execution Payload
Apify LinkedIn & Web Scraper
Search Parameters → Raw Prospect Profile JSON
Key Engineering Decisions
Each AI module has a single responsibility, making the system easier to test, maintain, replace, and improve independently.
Implemented a mandatory 'Human Review' checkpoint before email dispatch to guarantee campaign safety.
Decoupled workflow orchestration (n8n) from core application APIs (FastAPI) to handle long-running async tasks smoothly.
Technology Stack
Production Technologies
Frontend dashboard for campaign management and human review interface
Utility-first styling for the campaign management UI
Python backend exposing AI module APIs and webhook handlers
GPT-4 powering qualification, website analysis, and email generation modules
Lead store, campaign state, and review queue with real-time subscriptions
Web scraping and company data discovery at scale
Workflow orchestration for multi-step pipeline with independent step retries
Containerized deployment ensuring consistent environments
Features & Capabilities
What It Does
Lead Discovery
Apify-powered scraping finds and stores qualified company prospects automatically.
ICP Qualification
AI scores each lead against your ideal customer profile before spending tokens on outreach.
Website Intelligence
Scrapes and synthesizes company sites into actionable sales context.
Personalized Emails
AI generates tailored cold emails that reference specific company context.
Human Review Gate
Mandatory review checkpoint — no email is ever sent without human approval.
Follow-Up Sequences
Day-based automated follow-up generation (Day 0, 3, 7) without manual scheduling.
Reply Classification
AI reads and categorizes incoming replies into actionable intent buckets.
Campaign Analytics
Pipeline performance, open rates, and conversion data in real-time dashboard.
Workflow Pipeline
Step-by-Step System Flow
Lead Discovery
Apify scraper finds companies matching target criteria and populates Supabase lead store.
ICP Qualification
AI qualification module scores each lead against ideal customer profile criteria.
Website Analysis
Company website is scraped and analyzed to extract key sales vectors and pain points.
Email Generation
Personalized first-touch email drafted based on website analysis and prospect persona.
Human Review
All generated emails shown to human reviewer before any outreach is sent.
Outreach Dispatch
Approved emails sent via email API. Campaign tracking begins.
Follow-Up Sequences
Day 0, Day 3, Day 7 follow-up emails generated and dispatched automatically.
Reply Classification
Incoming replies analyzed and categorized: Interested, Not Now, Not Interested, Unclear.
Campaign Analytics
Pipeline metrics, conversion rates, and recommendations displayed in dashboard.
Interface & Dashboard
Alpha Lead Engine Interface
Analytics Dashboard
Conversation View
Engineering Challenges
Hard Problems Solved
The Problem
Generic cold emails get ignored. But writing truly personalized emails for hundreds of prospects manually is not viable.
Engineering Solution
Built a two-step pipeline: first scrape and vectorize the company website, then use that context as RAG input to the email generation module. Every email references something real about the specific company.
Results & Evaluation
What Was Achieved
Future Roadmap
What's Next
7 planned features · Active development
Source Code
GitHub Repository & Codebase
A live outbound sales automation platform that discovers leads, qualifies them with AI, generates personalized outreach emails, automates multi-step follow-ups, and classifies replies — reducing manual prospecting to a single reviewable pipeline.