> ## Documentation Index
> Fetch the complete documentation index at: https://docs.useorbitcrm.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# AI SDR Qualifying

> Automatically classify inbound leads into qualification buckets using AI. Get human-readable reasoning and recommended next actions for every submission.

## Overview

The AI SDR Qualifying node acts like an intelligent sales development rep. Instead of using rigid rules or point-based scoring, it analyzes each lead holistically and classifies them into meaningful qualification buckets with clear reasoning.

### What makes it different

> [card] Classification, not scoring
>
> Assigns leads to qualification buckets instead of numeric scores. Sales knows exactly what action to take.

> [card] Human-readable reasoning
>
> Every classification comes with clear explanations your team can understand and trust.

## How to Use

**1. Add the SDR Agent node**

In your workflow, click "Add Node" and select "SDR Qualifying" from the AI Processing category.

**2. Connect it after enrichment (recommended)**

For best results, place the SDR Agent after a Clay enrichment node so the AI has company data to analyze.

**3. Define your ICP**

Write a description of your ideal customer profile. Be specific about company size, industry, and characteristics you're looking for.

**4. Add disqualify rules (optional)**

Add field-based auto-disqualify rules (e.g., "Email contains @gmail.com" or "Company is empty"). Leads matching a rule are assigned to a non-qualified bucket immediately, skipping the AI call entirely — this saves AI credits on obvious non-fits.

**5. Use the outputs downstream**

Access classification results via template variables like `{{agent.bucket}}`, `{{agent.primary_reasons}}`, and `{{agent.recommended_action}}`.

## Configuration

### ICP Description (Required)

A free-text description of your ideal customer profile. The AI uses this to evaluate fit.

```
Example: "B2B SaaS companies doing $1M–$50M ARR with internal marketing teams that actively run paid ads."
```

### Auto-Disqualify Rules (Optional)

Deterministic field rules that run before the AI. Each rule combines a field, an operator (equals, contains, is empty, etc.), and a value. A lead matching any rule is classified into a non-qualified bucket without an AI call, so you don't spend AI credits on obvious non-fits.

Example: `Email` `contains` `@gmail.com` if you only want business emails, or `Company` `is empty` to skip leads with no company.

## Qualification Buckets

Every agent starts with two default buckets — **Qualified** and **Not Qualified** — and you can add, rename, describe, or remove buckets to match your process (a minimum of two is required). Optional templates like **High-Intent ICP Match**, **Good ICP, Low/Medium Intent**, **Potential Future Fit**, **Wrong ICP**, and **Spam / Invalid** can be added for more granular routing.

For each bucket you control:

| Setting | Description |
| - | - |
| **Name & description** | What the bucket is called and what makes a lead fit it — the description guides the AI's classification |
| **Qualified toggle** | Whether leads in this bucket count as qualified for routing and filtering |
| **Contact status** | Optionally write a qualification status (Qualified, Unqualified, Pending) back to the contact record when a lead lands in this bucket |

The AI classifies each lead into exactly one of your buckets with a confidence score and reasoning.

## Output Variables

| Variable | Description |
| - | - |
| `{{agent.bucket}}` | The assigned qualification bucket |
| `{{agent.confidence}}` | Confidence score from 0.0 to 1.0 |
| `{{agent.primary_reasons}}` | Array of reasons explaining the classification |
| `{{agent.risks_or_unknowns}}` | Any uncertainties or missing information |
| `{{agent.recommended_action}}` | Suggested next step |

## Best Practices

* Be specific in your ICP description — the more context, the better the AI can classify leads
* Use enrichment data from Clay before the SDR Agent for richer classification
* Route high-intent leads to immediate notifications (Slack, email) and CRM creation
* Use a Filter node after SDR Agent to branch based on bucket (e.g., only send qualified leads to HubSpot)
* Review classification results periodically to refine your ICP description

## Example Workflow

A typical lead qualification workflow using the SDR Agent:

**Flow:** Form Submission → Clay Enrichment → SDR Agent → Router

**Router conditions:**

* Qualified → Slack notification + HubSpot
* Not Qualified → No action (or a long-term nurture sequence)

If you've added more granular buckets (e.g. High-Intent ICP Match, Potential Future Fit), you can route each one to a different path.

## Next Steps

> [card] [AI SDR Summarizer](https://docs.useorbitcrm.ai/guides/ai-nodes/summarizer.md)
>
> Generate AI summaries of submissions

> [card] [All Node Types](https://docs.useorbitcrm.ai/guides/workflows/nodes.md)
>
> Explore all workflow nodes


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.