Resumen
Activar on nuevo CRM contact, enriquecer con un Scavio Google search for company context, puntuacion con un short LLM prompt (3 reglas), y etiqueta hot/warm/cold in el CRM. Total: 12 lines of expression code a traves de 4 n8n nodes.
Desencadenador
Nuevo contact created in CRM (webhook o polling)
Programación
On nuevo CRM contact
Pasos del flujo de trabajo
Webhook activar on nuevo CRM contact
n8n recibe company nombre + correo electronico de CRM webhook.
Scavio HTTP Solicitud node
POST to /api/v1/search con plataforma=google, consulta=company nombre. Returns structured SERP.
LLM scoring node
Enviar top-3 SERP fragmentos + company nombre to Claude/GPT con prompt: 'Puntuacion 1-10: hiring senales, funding news, tech stack fit. Return JSON {puntuacion, reason}.'
IF node: route by puntuacion
puntuacion >= 7 → hot (assign to AE), 4-6 → warm (nurture sequence), <4 → cold (archive).
Implementacion en Python
import requests, os
key = os.environ["SCAVIO_API_KEY"]
company = "Acme Corp"
resp = requests.post("https://api.scavio.dev/api/v1/search",
headers={"x-api-key": key},
json={"query": company, "platform": "google", "limit": 3})
snippets = [r["snippet"] for r in resp.json().get("results", [])]
prompt = f"Score this company 1-10 for B2B SaaS fit: {company}. Context: {snippets}. Return JSON: score, reason."
score_result = call_llm(prompt)
print(score_result)Implementacion en JavaScript
const resp = await fetch("https://api.scavio.dev/api/v1/search", {
method: "POST",
headers: { "x-api-key": process.env.SCAVIO_API_KEY, "Content-Type": "application/json" },
body: JSON.stringify({ query: "Acme Corp", platform: "google", limit: 3 })
});
const snippets = (await resp.json()).results.map(r => r.snippet);
const score = await callLLM(`Score 1-10 for B2B SaaS fit: Acme Corp. Context: ${snippets.join(" ")}. Return JSON: {score, reason}.`);
console.log(score);Plataformas utilizadas
Búsqueda web con grafo de conocimiento, PAA y resúmenes de IA