An n8n workflow that classified and scored inbound leads, routed qualified prospects, and reduced daily manual data entry.
Inbound lead data arrived incomplete and required repeated human review before it could be routed into a CRM pipeline.
Designed the workflow, AI classification boundary, scoring and routing logic, HubSpot integration, PostgreSQL state, error paths, and handover.
AI outputs had to be converted into controlled business decisions while maintaining traceable state and a path for uncertain cases.
Client identity and lead sources are confidential.
Prompts, qualification rules, and volumes are withheld.
The case does not claim fully autonomous sales decisions.
The workflow normalized lead data, enriched context, called model providers behind explicit steps, applied scoring and routing rules, and persisted state before CRM updates.
Explicit routing branches
Persistent lead state
Manual-review path for uncertainty
Private prompts and business rules withheld
The workflow reduced daily manual data entry and routed qualified prospects, according to the supplied resume. No unsupported lead volume or revenue figure is claimed.
Tayyab can combine n8n, AI, databases, and CRM APIs without hiding operational decisions inside a single prompt.
AI output should feed explicit routing rules, not replace them.
Uncertain cases need a human path.
Persist state before making external CRM changes.