background Layer 1

AI-Powered Intelligent SDR: Autonomous B2B Lead Qualification

Customer
Total Express
Project manager on the customer side
Juliano De Conti
CIO
Year of project completion
2026
Project timeline
october, 2025 — september, 2026
Project scope
7392 man-hours
Goals

Turn B2B lead qualification at one of Brazil's largest logistics and e-commerce delivery companies from a manual, rep-dependent process into an AI-driven, always-on digital operation. Specific goals: (1) shorten the lead qualification cycle, which averaged 38 days; (2) reduce the cost of qualifying each lead; (3) standardize qualification through an objective, automated Lead Score; (4) eliminate manual CRM updates and keep a complete, traceable history of every interaction; (5) give B2B prospects an immediate 24/7 response on the channel they already use, WhatsApp; and (6) let account executives focus only on sales-qualified opportunities, so the commercial pipeline can scale without proportional headcount growth.

Project Results

Results. Qualification cycle cut from 38 days to about one week (-80%), with same-day proposals for engaged leads. Cost per qualification down from R$ 56.86 to R$ 11.57 (-80%, 4.9x cheaper). 24/7 availability, standardized qualification, automatic CRM updates with full traceability, and a sales team focused on qualified opportunities.

The uniqueness of the project

This is not a chatbot or a scripted funnel. The AI works as an autonomous pre-sales agent that does a human SDR's job end to end. It opens the conversation on WhatsApp, understands free-text questions, answers from a corporate knowledge base, handles objections and runs the qualification interview. In the same conversation it validates the prospect against official business registries (tax ID, registration status, segment, address, company size), computes a Lead Score and decides on its own whether the lead stays in nurturing or becomes an SQL. It then books the meeting in the account executive's calendar and writes everything back to the CRM. People step in only on hand-off. Result: qualification cycle cut from 38 days to about one week, cost per qualified lead down ~80% (4.9x cheaper), and same-day proposals for engaged leads.

Used software

• n8n – workflow orchestration of the full lead journey

• Meta WhatsApp Business Cloud API – official WhatsApp messaging channel

• Chatwoot – conversation hub, centralized history and human hand-off

• Salesforce – corporate CRM (automatic lead and opportunity updates)

• CNPJá – Brazilian company registry API for data enrichment and validation

• AI engine – large language model grounded in a corporate knowledge base [confirm model/provider]

• Corporate calendar – automatic meeting scheduling with account executives

Difficulty of implementation

The challenge was letting an AI make commercial decisions inside a live conversation without losing control or quality. That required: (1) orchestrating six systems (messaging channel, conversation hub, CRM, registry API, AI engine and calendar) so every message, validation and status change stays consistent and traceable; (2) turning tacit sales judgment into explicit, auditable Lead Score rules; (3) running registry validations mid-conversation while keeping it natural; (4) grounding answers in a curated knowledge base so the agent stays accurate and on-brand; (5) designing hand-off rules so a person takes over when the AI should not proceed; and (6) handling personal and company data in line with Brazil's data protection law (LGPD) and WhatsApp Business messaging policies. Adoption was the other half: sales teams had to trust AI-qualified leads and rely on the automated CRM.

Project Description

Context. Total Express is one of Brazil's largest logistics and e-commerce delivery companies. Growth depends on onboarding new B2B shippers quickly, and the pre-sales step had become a bottleneck.

Challenge. Lead qualification relied on manual pre-sales work: a long delay between lead capture and first contact (the qualification cycle averaged 38 days), heavy repetitive workload, no standard qualification criteria, fragmented interaction history, manual CRM updates and opportunities lost to slow response. The process could not scale with demand.

Solution. We built an Intelligent SDR: a digital pre-sales agent made of specialized AI agents and integrated with our core commercial systems. It automates the whole journey: Campaigns → MQL → Intelligent SDR → Lead Score → SQL → Meeting → Account Executive.

1. Intake. Campaign leads are received automatically and classified as MQLs.

2. Conversational AI. The agent contacts the lead on WhatsApp, understands natural language, answers from a corporate knowledge base, identifies intent, handles objections and adapts to the lead's profile.

3. Real-time validation. During the conversation it checks the company's tax ID (CNPJ), registration status, economic segment, address and size through a business-registry API.

4. Scoring and decision. Business rules combine company data and stated needs into a Lead Score; the agent decides whether to keep nurturing the lead or promote it to SQL.

5. Hand-off. For SQLs, the meeting is booked automatically in the account executive's calendar, the CRM is updated, the full conversation is kept in the service hub, and a person can take over at any time.

Architecture. API-oriented and scalable: n8n orchestrates the workflows; the WhatsApp Business Cloud API provides the official channel; Chatwoot centralizes conversations; Salesforce is the CRM; CNPJá enriches company data; an AI engine grounded in our knowledge base drives the dialogue.

Results. Qualification cycle cut from 38 days to about one week (-80%), with same-day proposals for engaged leads. Cost per qualification down from R$ 56.86 to R$ 11.57 (-80%, 4.9x cheaper). 24/7 availability, standardized qualification, automatic CRM updates with full traceability, and a sales team focused on qualified opportunities.

Next steps. Predictive lead-scoring models, sentiment analysis, next-best-offer recommendations, real-time executive dashboards and AI-generated commercial proposals.

Project geography

Brazil (nationwide)

We use cookies for analytical purposes and to deliver you the best experience with our website. Continuing to the site, you agree to the Cookie Policy.