How Email Intent Classifier Automates Intent Scoring

AI reads inbound emails, scores buyer intent across 7 categories, and routes to Pipedrive. That single sentence captures a workflow gap that costs sales, revops teams hours every week. The manual process behind what Email Intent Classifier automates is familiar to anyone who has worked in a revenue organization: someone pulls data from Pipedrive, Gmail, copies it into a spreadsheet or CRM, applies a mental checklist, writes a summary, and routes it to the next person in the chain. Repeat for every record. Every day.

Three problems make this unsustainable at scale. First, the process does not scale. As volume grows, the human bottleneck becomes the constraint. Whether it is inbound leads, deal updates, or meeting prep, a person can only process a finite number of records before quality degrades. Second, the process is inconsistent. Different team members apply different criteria, use different formats, and make different judgment calls. There is no single standard of quality, and the output varies from person to person and day to day. Third, the process is slow. By the time a manual review is complete, the window for action may have already closed. Deals move, contacts change roles, and buying signals decay.

These are not theoretical concerns. They are the operational reality for sales, revops teams handling intent scoring and revenue operations workflows. Every hour spent on manual data processing is an hour not spent on the work that actually moves the needle: building relationships, closing deals, and driving strategy.

This is the gap Email Intent Classifier fills.

Four Agents. Seven Intents. Emails Routed to Pipedrive Automatically.

Email Intent Classifier is a 30-node n8n workflow with 4 specialized agents. Each agent handles a distinct phase of the pipeline, and the handoff between agents is deterministic — no ambiguous routing, no dropped records. The blueprint is designed so that each agent does one thing well, and the overall pipeline produces a consistent, auditable output on every run.

Here is what each agent does:

• The Researcher (Tier 1 Reasoning): Receives inbound email data from Gmail and conducts web-based sender research — company intelligence, role seniority, and deal context from Pipedrive.
• The Classifier (Tier 1 Reasoning): Scores each email across 7 intent categories: DEMO_REQUEST, PRICING_INQUIRY, OBJECTION, RENEWAL_INTEREST, CHURN_SIGNAL, GENERAL_INQUIRY, and SPAM_OR_NOISE.
• The Router (Conditional): Three-way confidence gate: high confidence (≥ 0.7) routes to Pipedrive CRM — deals, activities, and notes created based on intent category.
• The Syncer (Integration): Writes to Pipedrive based on intent: DEMO_REQUEST, PRICING_INQUIRY, and RENEWAL_INTEREST create Deals + Activities.

When the pipeline completes, you get structured output that is ready to act on. The blueprint bundle includes everything needed to deploy, configure, and customize the workflow. Specifically, you receive:

• 30-node n8n workflow (.json) — you own it
• 2 production-ready agent system prompts
• 8 SDC inter-agent schemas
• Intent taxonomy with 7 categories and examples
• Confidence threshold tuning guide
• Error handling matrix (26 failure modes documented)
• Dependency matrix with ITP-measured costs
• README setup guide (15 minutes)

Every component is designed to be modified. The agent prompts are plain text files you can edit. The workflow nodes can be rearranged or extended. The scoring criteria, output formats, and routing logic are all exposed as configurable parameters — not buried in application code. This means Email Intent Classifier adapts to your specific process, terminology, and integration requirements without forking the entire workflow.

Understanding how the pipeline works helps you customize it for your environment and troubleshoot issues when they arise. Here is a step-by-step walkthrough of the Email Intent Classifier execution flow.

Step 1: The Researcher

Tier: Tier 1 Reasoning

Receives inbound email data from Gmail and conducts web-based sender research — company intelligence, role seniority, and deal context from Pipedrive. Outputs a structured research payload that gives the Classifier the context it needs for accurate intent scoring.

This stage is critical because it ensures that downstream agents receive structured, validated input. Each agent in the pipeline trusts the output contract of the previous agent. If The Researcher identifies an issue — a missing field, a low-confidence score, or an unexpected input format — the pipeline handles it explicitly rather than passing garbage downstream. This is the difference between a prototype and a production-grade workflow: every handoff is defined, every edge case is documented.

Step 2: The Classifier

Tier: Tier 1 Reasoning

Scores each email across 7 intent categories: DEMO_REQUEST, PRICING_INQUIRY, OBJECTION, RENEWAL_INTEREST, CHURN_SIGNAL, GENERAL_INQUIRY, and SPAM_OR_NOISE. Outputs intent classification with a confidence score (0–1.0) and chain-of-thought reasoning.

This stage is critical because it ensures that downstream agents receive structured, validated input. Each agent in the pipeline trusts the output contract of the previous agent. If The Classifier identifies an issue — a missing field, a low-confidence score, or an unexpected input format — the pipeline handles it explicitly rather than passing garbage downstream. This is the difference between a prototype and a production-grade workflow: every handoff is defined, every edge case is documented.

Step 3: The Router

Tier: Conditional

Three-way confidence gate: high confidence (≥ 0.7) routes to Pipedrive CRM — deals, activities, and notes created based on intent category. Low confidence routes to human review with full classification context. SPAM/noise logs and skips with zero additional LLM calls.

This stage is critical because it ensures that downstream agents receive structured, validated input. Each agent in the pipeline trusts the output contract of the previous agent. If The Router identifies an issue — a missing field, a low-confidence score, or an unexpected input format — the pipeline handles it explicitly rather than passing garbage downstream. This is the difference between a prototype and a production-grade workflow: every handoff is defined, every edge case is documented.

Step 4: The Syncer

Tier: Integration

Writes to Pipedrive based on intent: DEMO_REQUEST, PRICING_INQUIRY, and RENEWAL_INTEREST create Deals + Activities. OBJECTION and CHURN_SIGNAL create Notes on existing deals. GENERAL_INQUIRY creates an Activity. Non-blocking writes — pipeline never stalls on CRM errors.

This stage is critical because it ensures that downstream agents receive structured, validated input. Each agent in the pipeline trusts the output contract of the previous agent. If The Syncer identifies an issue — a missing field, a low-confidence score, or an unexpected input format — the pipeline handles it explicitly rather than passing garbage downstream. This is the difference between a prototype and a production-grade workflow: every handoff is defined, every edge case is documented.

The entire pipeline executes without manual intervention. From trigger to output, every decision point is deterministic: if a condition is met, the next agent fires; if not, the record is handled according to a documented fallback path. There are no silent failures. Every execution produces a traceable audit trail that you can review, export, or feed into your own reporting tools.

This architecture follows the ForgeWorkflows principle of tested, measured, documented automation. Every node in the pipeline has been validated during ITP (Inspection and Test Plan) testing, and the error handling matrix in the bundle documents the recovery path for each failure mode.

All values below are from ITP testing — not estimates, not projections. Measured across 20 test emails spanning all 7 intent categories, borderline confidence cases, and edge cases.

The primary operating cost for Email Intent Classifier is the per-execution LLM inference cost. Based on ITP testing, the measured cost is: Cost per Email (ITP-Measured): $0.116/email (ITP-measured blended avg). This figure includes all API calls across all agents in the pipeline — not just the primary reasoning step, but every classification, scoring, and output generation call.

To put this in context, consider the manual alternative. A skilled team member performing the same work manually costs $50–75/hour at a fully loaded rate (salary, benefits, tools, overhead). If the manual version of this workflow takes 20–40 minutes per cycle, that is $17–50 per execution in human labor. The blueprint executes the same pipeline for a fraction of that cost, with consistent quality and zero fatigue degradation.

Infrastructure costs are separate from per-execution LLM costs. You will need an n8n instance (self-hosted or cloud) and active accounts for the integrated services. The estimated monthly infrastructure cost is $10–60/month depending on email volume, depending on your usage volume and plan tiers.

Quality assurance: BQS audit result is 12/12 PASS. ITP result is 14/14 PASS, 20/20 intent accuracy (100%). These are not marketing claims — they are test results from structured inspection protocols that you can review in the product documentation.

9 files. Workflow, 2 system prompts, intent taxonomy, threshold guide, and complete documentation.

When you purchase Email Intent Classifier, you receive a complete deployment bundle. This is not a SaaS subscription or a hosted service — it is a set of files that you own and run on your own infrastructure. Here is what is included:

• email_intent_classifier_v1_0_0.json — The 30-node n8n workflow (4-agent pipeline with 3-way routing)
• README.md — Setup guide (15 minutes)
• CHANGELOG.md — Version history
• system_prompt_researcher.txt — Researcher agent system prompt (sender intelligence)
• system_prompt_classifier.txt — Classifier agent system prompt (7-class intent scoring)
• intent_taxonomy.md — Complete intent category definitions and examples
• confidence_threshold_guide.md — Threshold tuning guide for your email volume
• blueprint_dependency_matrix.md — Required services, API keys, ITP-measured costs
• itp_results.md — Full test results: 20 emails, 14/14 milestones PASS

Start with the README.md. It walks through the deployment process step by step, from importing the workflow JSON into n8n to configuring credentials and running your first test execution. The dependency matrix lists every required service, API key, and estimated cost so you know exactly what you need before you start.

Every file in the bundle is designed to be read, understood, and modified. There is no obfuscated code, no compiled binaries, and no phone-home telemetry. You get the source, you own the source, and you control the execution environment.

Email Intent Classifier is built for Sales, Revops teams that need to automate a specific workflow without building from scratch. If your team matches the following profile, this blueprint is designed for you:

• You operate in a sales or revops function and handle the workflow this blueprint automates on a recurring basis
• You have (or are willing to set up) an n8n instance — self-hosted or cloud
• You have active accounts for the required integrations: Pipedrive CRM, Gmail account
• You have API credentials available: Anthropic API, Pipedrive API, Gmail OAuth2
• You are comfortable importing a workflow JSON and configuring API keys (the README guides you, but basic technical comfort is expected)

This is NOT for you if:

• Does not send reply emails — output is Pipedrive records and classification data
• Does not process email attachments — classification is based on subject and body text only
• Does not support CRMs other than Pipedrive in v1 — Pipedrive-specific deal, activity, and note creation
• Does not handle email threading — each inbound email is classified independently

Review the dependency matrix and prerequisites before purchasing. If you are unsure whether your environment meets the requirements, contact support@forgeworkflows.com before buying.

Deployment follows a structured sequence. The Email Intent Classifier bundle is designed for the following tools: n8n, Anthropic API, Pipedrive, Gmail. Here is the recommended deployment path:

1. Step 1: Connect Gmail and configure credentials. Import the n8n workflow JSON and configure your Gmail OAuth2, Anthropic API key, and Pipedrive API token.
2. Step 2: Review intent taxonomy and adjust threshold. Review the 7 intent categories in the included taxonomy guide. Adjust the confidence threshold (default 0.7) based on your email volume and review capacity.
3. Step 3: Activate and receive classified emails. Activate the workflow. Inbound emails are automatically researched, classified by intent, and routed — high confidence to Pipedrive, low confidence to human review, spam to log.

Before running the pipeline on live data, execute a manual test run with sample input. This validates that all credentials are configured correctly, all API endpoints are reachable, and the output format matches your expectations. The README includes test data examples for this purpose.

Once the test run passes, you can configure the trigger for production use (scheduled, webhook, or event-driven — depending on the blueprint design). Monitor the first few production runs to confirm the pipeline handles real-world data as expected, then let it run.

For technical background on how ForgeWorkflows blueprints are built and tested, see the Blueprint Quality Standard (BQS) methodology and the Inspection and Test Plan (ITP) framework. These documents describe the quality gates every blueprint passes before listing.

Ready to deploy? View the Email Intent Classifier product page for full specifications, pricing, and purchase.