💡 Executive SummaryProject Type: Work Automation & AI Integration Tech Stack: Make.com, Custom Webhooks, Postman, Custom HTTP API, Google Gemini AI, Slack, Google Sheets Developer: Dimas Dwi Antoko | Automation Engineer Deliverable: Enterprise-Grade Fault-Tolerant Lead Processing System

📌 1. The Business Problem

B2B SaaS and agency sales teams spend over 15 hours per week manually reviewing unqualified lead forms, searching company data, and typing updates into CRMs.

• Response Latency: High-intent leads wait hours for a reply, dropping conversion rates.
• Manual Data Entry Error: Incomplete profiles lead to poor sales outreach context.
• No Real-Time Alerting: Sales teams miss urgent $10k+ opportunities because form submissions sit in email inboxes.

🛠️ 2. The Automated Solution & Architecture

An automated, end-to-end processing pipeline built in Make.com. Inbound leads are validated via Webhooks, enriched with domain data using a Custom HTTP API, scored (1–100) by Google Gemini AI, and dynamically routed to Slack or Google Sheets.

Scenario Flow Architecture

[Custom Webhook] ➔ [HTTP API Enrichment] ➔ [Google Gemini AI] ➔ [Parse JSON] ➔ [Flow Control Router]

• Branch 1 (Hot Leads $\ge$ 70): [Slack Notification]
• Branch 2 (Nurturing Leads < 70): [Google Sheets Database]

⚙️ 3. Technical Features & Implementation Details

• Real-Time Webhook Trigger: Built a custom JSON endpoint accepting inbound payloads tested with Postman.
• Dynamic Domain Extraction: Used native text functions split() to extract target company domains from email inputs.
• Structured AI Lead Scoring: Configured Google Gemini AI to analyze inquiry text, company size, and domain data, outputting schema-strict JSON without markdown wrapping.
• Automated Data Normalization: Implemented a JSON Parser module to break raw AI outputs into actionable execution parameters (score, qualification, reason).