Prospector

🇪🇸 ES Only

Prospector turns the official Spanish business registry into a living pipeline of qualified B2B leads. Every day it mines the BORME (Boletín Oficial del Registro Mercantil) for strategic moves, enriches each one with real contact, web and technology data gathered from open sources (OSINT), and scores it as a potential buyer of whatever you sell — all running on-premise, on your own infrastructure.

🇪🇸 Spain Only

Prospector works exclusively with companies registered in Spain. Its entire pipeline is built around the BORME and Spanish business data sources — it does not cover companies from other countries. If your target market is Spain, this module is built for you.

Features

• AI Scoring: every lead is scored by AI as a potential buyer, so you always know who to call first
• Product / Company Matching: each company is matched against your catalog for real product-market fit — not a generic score
• Add Multiple of Your Products: build your own sales catalog with as many of your products as you want; every lead is analyzed and pitched against your actual offering
• OpenAI or Local GPT: works with your OpenAI API key or fully on-premise with Local GPT — your choice of cloud or 100% private
• Built-in CRM: contacted / discarded / favorites, contact history, notes and follow-up reminders, all inside HomeDock OS
• People & Staff Tracking: keep track of the key people and decision-makers behind each company
• Direct Apollo & LinkedIn Links: jump straight to Apollo and LinkedIn for each lead to supercharge your prospecting
• OSINT Enrichment: phone, email, website, CNAE and tech stack gathered from 9+ open, publicly available Spanish data sources
• Validated Contact Data: phones checked in parallel, emails verified via MX + cascade fallback
• AI Outreach: fit analysis, company profiling, first-contact message and follow-up — all written for you
• Daily BORME Mining: automatically discovers companies making strategic moves across the provinces you choose
• Auto-Prospector Bot: a background daemon that enriches and analyzes new leads automatically, around the clock

Why Prospector

• Buying-intent signal, not a cold list: strategic moves in the registry are one of the strongest early indicators that a company is growing and ready to invest.
• Tailored to what you sell: configure your sales catalog once, and every lead is scored, analyzed and pitched against your own products and methodology.
• Fully on-premise: scraping, enrichment and AI can all run locally with Local GPT — no lead data ever has to leave your network.
• A CRM built in: mark, filter, favorite and follow up on every lead without leaving HomeDock OS.

How It Works

Prospector runs a four-stage pipeline that takes a company from raw registry entry to sales-ready lead.

1. Scrape

Pick a date and the provinces you want to target. Prospector downloads that day’s BORME PDFs, extracts the text, and parses every relevant entry — company name, amount, address, business purpose and directors — saving them all to your local history.

2. Enrich

For each company, Prospector runs an OSINT sweep across multiple open, publicly available data sources to find phone, email, website, CNAE and the technology the company uses. Everything comes from open, publicly accessible data. Phones are validated in parallel and emails are verified through MX checks with a cascading fallback, so what lands in your pipeline is data you can actually use.

3. AI Analysis

Once a company is enriched, four AI actions bring it to life — powered by OpenAI or, for full privacy, Local GPT running on-premise:

• Analyze: product fit, urgency/regulation signals, target role and headcount
• X-Ray: a concise profile and estimation of the company
• Prospect: a ready-to-send first-contact message or email
• Follow-up: a tailored follow-up message to keep the conversation going

_{Every AI action is scored against your own sales catalog and B2B discovery methodology, configured on first launch.}

Great Results on Modest Hardware

Running fully on-premise? The Phi-3.5 Mini 3.8B model available in Local GPT, along with the different Qwen models, delivers solid results locally even on machines with limited resources — so you can prospect with full privacy without needing a powerful server.

4. CRM

Work your leads without ever leaving HomeDock OS. Mark companies as Contacted (with full contact history and follow-up intervals), Discarded (with a reason), or Favorite (starred). Add notes and set follow-up reminders so no opportunity slips through.

Auto-Prospector

Prospector ships with a background bot that does the busywork for you. On a configurable interval it automatically enriches and analyzes pending leads — enriching, analyzing, profiling and even drafting outreach — so you wake up to a pipeline that already qualified itself overnight.

Lightweight by Design

The Auto-Prospector runs as a separate background worker with a minimal footprint — it barely consumes resources while it quietly picks up new leads as they arrive.

Lead Filtering

The Leads view gives you a fast, focused way to work your pipeline:

• Filters: All / Pending / Contacted / Follow-up / Discarded / Favorites
• Sorting: by Score or by Activity
• Search: find any company instantly

An icon-only design keeps the interface clean, with labels revealed on hover.

Configuration

• AI Provider: OpenAI (API key + model, default gpt-5-nano) or Local GPT (fully on-premise)
• Default Provinces: Murcia, Madrid, Barcelona and Valencia — fully customizable
• CNAE Exclusion: a CNAE-2009 tree filter that auto-discards off-target sectors after enrichment, before AI
• Sales Catalog: your products and B2B methodology, configured on first launch and used across every AI action

Configure Once, Sell Everywhere

The vendor catalog is the heart of Prospector. Describe what you sell and how you sell it once, and every lead is scored, analyzed and pitched against your offering — automatically.

Technical Details

• Data Source: BORME (official Spanish Mercantile Registry)
• Enrichment: OSINT sweep across multiple open, publicly available data sources
• AI Engine: OpenAI API or Local GPT (on-premise)
• Streaming: Server-Sent Events (SSE)