VC Lead Enrichment Automation System

Didi

Didi Berman

AI-Powered Investor Intelligence Engine for Capital Introductions

A fully automated lead enrichment system designed specifically for a UK-based capital introductions firm. It transforms basic lead information from Zoho CRM into a rich, actionable investor profile by leveraging the power of n8n.io and a multi-agent Artificial Intelligence framework.

The Challenge: From Raw Data to Actionable Intelligence

In the fast-paced world of capital introductions, speed and accuracy are paramount. Manually researching new leads (investors) is a time-consuming bottleneck, prone to inconsistencies. The goal was to eliminate this manual work and provide the team with deep, relevant insights directly within their CRM the moment a new lead is captured.

The Solution: A Resilient, Multi-AI Automation Workflow

This system is built on n8n, a workflow automation platform, to orchestrate a seamless process from data capture to final enrichment.
Here's how it works:
Instant Trigger: The workflow is initiated instantly via a webhook from Zoho CRM whenever a new lead is created.
Intelligent Deduplication: To ensure efficiency and prevent redundant processing, the system first checks an internal database to see if the lead has been researched before. If it's a duplicate, the process stops, saving valuable resources.
The AI Gauntlet - A "Best-of-Breed" Approach: This is where the magic happens. Instead of relying on a single AI, the system queries three different state-of-the-art Large Language Models (LLMs) in a cascading sequence to find the best possible information:
Round 1 (Perplexity): The lead data is first sent to Perplexity's Sonar Pro model with a highly detailed prompt, requesting key investor data points like investment sectors, typical ticket size, investment criteria, and a concise company summary.
Round 2 (Anthropic Claude - Conditional Fallback): If the first model fails to find crucial details (like the employee's specific role), the system automatically triggers a second, more focused search using Anthropic's Claude 3.5 Sonnet.
Round 3 (Google Gemini - The Final Check): If both previous models come up short, the system makes a final attempt with Google's Gemini 1.5 Pro.
The AI Judge: A custom scoring algorithm analyzes the results from all three AIs. It rates each response based on completeness and data quality (e.g., does the role contain a valid job title? Is the sector focus detailed?). The highest-scoring response is selected as the definitive source of truth. If all AIs perform poorly, the system will even merge the best available data points from each to create a composite profile.
Seamless CRM Update: The winning data is meticulously parsed and used to automatically update the lead's record in Zoho CRM. Custom fields for "Investment Sectors," "Company Summary," "Ticket Size," and more are populated instantly, turning a basic contact into a fully qualified investor profile.

Key Features & Benefits

Zero Manual Research: Frees up the capital introductions team to focus on building relationships instead of tedious data entry.
Deep Investor Insights: Goes beyond contact info to provide the critical data needed to qualify and engage potential investors.
Unmatched Accuracy & Resilience: The multi-AI fallback and scoring system dramatically increases the probability of finding high-quality, accurate information.
Cost-Efficient Operation: Smart deduplication and a cascading AI approach prevent unnecessary API calls and reduce operational costs.
Fully Integrated: Operates as a silent, background-level enhancement to the team's existing Zoho CRM workflow.

Technology Stack

Automation Platform: n8n.io
CRM: Zoho CRM
AI Orchestration: LangChain
AI Models: Perplexity (Sonar Pro), Anthropic (Claude 3.5 Sonnet), and Google (Gemini 1.5 Pro) via OpenRouter.
Like this project

Posted Oct 7, 2025

Developed lead enrichment system integrating ZohoCRM with Custom AI APIs, through n8n.