FundEDU: A Fully Automated Lead Intelligence Pipeline for EITC Donor Outreach

The Challenge

Reps at FundEDU spent hours searching Google Maps for businesses with no automated pipeline or density control, then 15 to 30 minutes researching each business before calls with no structured AI support. As the business scaled, outreach operations became entirely reliant on manual research with no scalable system in place.

FundEDU operates in Pennsylvania's EITC program, a competitive, relationship-driven space where donor outreach is the core revenue engine. The business depends on consistently identifying, qualifying, and reaching the right businesses at the right time with the right message.

What Operating in EITC Outreach Means

• High lead volume with no structured pipeline
• Multiple schools, regions, and donor eligibility windows
• Time-sensitive outreach and strict qualification requirements
• Growth that was impossible without proportional research headcount

Key Challenges Faced

Manual Lead Discovery

Reps spent hours searching Google Maps for businesses with no automated pipeline or density control.

No Lead Prioritisation

No system existed to rank leads, so reps had no way to know who to call first or why.

Manual Pre-Call Research

Reps spent 15 to 30 minutes researching each business before calls with no structured AI support.

No Delivery History

Contacts were regularly re-approached without record of prior interaction.

Unverified Contact Data

Phone numbers and emails went unchecked, wasting rep time on dead contacts and bounced addresses.

The Solution

TekConnected built a fully automated lead intelligence pipeline. Instead of treating lead research as a manual rep task, it was rebuilt as an automated system.

The Stack

Google Places + PA State Datasets

Discovers businesses via Google Places 24/7 using a 1km squared tile grid, and ingests 7 PA state datasets automatically.

n8n Automation (19 Workflows)

Pulls structured data, validates records, resolves duplicates, scores leads, and delivers ranked lists to reps.

Supabase + OpenAI (AI Profiles)

Stores 280+ tables of verified business data and generates AI profiles per lead using the OpenAI API.

Four-Layer Data Pipeline Design

TekConnected designed the entire lead lifecycle across four stages, from Data Foundation through Canonicalization to Evaluation and Consolidation, identifying where research logic should be automated rather than manually executed.

Supabase as the Source of Truth

• All business data covering contact info, enrichment, and scores is stored in a single canonical database
• Lead scores are computed only when predefined quality conditions are met
• Score histories are maintained so the system detects when lead value changes

n8n as the Automation Control Layer

• Discovers, deduplicates, and verifies business records
• Generates AI intelligence profiles via the OpenAI API
• Delivers ranked, rep-ready leads to a shared Google Sheet
• Monitors for anomalies and dataset version changes

Leads are delivered automatically to the client's Google Sheet with verified contacts, AI profiles, and outreach scores, compatible with any CRM without changing existing tools. The underlying principle is that outreach intelligence should be system-driven, not rep-driven.

Technology Stack

Results & Outcomes

60 to 120 Minutes Per Day Saved Per Rep

On lead discovery alone.

3 Scores Per Lead

Donor potential, EITC eligibility, and outreach priority.

18 Workflows Running Continuously

In the background, with zero manual input.

100% Verified Contact Data

Before it reaches any rep.

Under 60 Seconds

For on-demand geographic search to deliver results.

Strategic Impact

• Reps start every day with a fresh, pre-ranked list of verified leads.
• Institutional knowledge of seven PA state datasets captured and continuously maintained.
• Foundation for full CRM replacement, eliminating approximately $12,000 per year in Salesforce licensing.
• Compounding intelligence: lead values re-evaluated automatically as conditions change.
• Scalable pipeline growth without proportional increase in research headcount.
• AI scheduling and Call Launchpad infrastructure already in place for future phases.

Who This Applies To

This build pattern fits an organisation that:

• Runs strong relationship networks alongside fragmented outreach operations
• Duplicates prospecting manually between Google Maps, spreadsheets, and call lists
• Gives reps no basis to know which lead to call first or why
• Cannot grow the pipeline without hiring proportional research headcount