A powersports financing company needed new lead acquisition channels. I identified an untapped marketplace, reverse-engineered its API, and built the entire pipeline from scraping to close.

A powersports financing company had the product but not the pipeline. Inbound alone wasn’t generating enough volume, and traditional outbound channels were either saturated or too expensive. They needed a creative acquisition channel that could find qualified prospects at scale without a sales team manually hunting for them.

Lead generation was limited to inbound traffic and manual outreach. No system for identifying prospects, no automated follow-up, no way to track what messaging was working. Every new lead required someone’s time to find, qualify, and engage.

I identified an untapped marketplace where their ideal prospects were already active and reverse-engineered its API to pull listing data without browser automation. From there I built a full pipeline.

A scraper grabs prospect data and enriches it through the marketplace’s API. Everything feeds into a custom web application I built in React — a dashboard where operators can view leads, respond, and track outreach performance. Auto-messaging rotates between active templates, and every template is tagged in the analytics so you can see exactly what’s converting and what’s not.

An AI agent lives in Discord and manages prospect responses autonomously. It has brand documents, interaction guidelines, and direct access to the application’s API so it can maintain context and execute tasks without human intervention. Every incoming message runs through sentiment analysis — high-intent responses trigger an instant Telegram notification so nothing slips through.

A background scraper continuously checks listing status through the marketplace API. When a listing is sold or removed and there’s no active conversation, the system cleans it out automatically. No wasted outreach to dead leads.

Alongside the pipeline, I built a landing page and email funnel to convert prospects once they engage.

The pipeline runs autonomously around the clock. Prospects get identified, contacted, and followed up with — all without manual effort. Operators check a dashboard, review high-intent notifications, and step in only when a prospect is ready to move forward. I built it, I operate it on a revenue share model, and I continue to improve it as the system surfaces new opportunities.

What started as a 4-week build has grown into an ongoing system that I continue to develop and operate. The pipeline has gone from initial concept to a fully autonomous lead generation engine — scraping, outreach, AI-managed follow-up, sentiment analysis, and listing hygiene all running without manual intervention. Multiple operators have been trained on the system and use it independently. The revenue share model means I have skin in the game — I only win when the pipeline performs.