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

# AI Scorer

> Quickly score every row in your dataset based on your own rules—so you always know which deals, leads, or accounts deserve the highest marks.

## What It Does

* **Custom Scoring:** Assign a numerical score to each row using your own plain-English rules.
* **Optional “Why?”:** Flip on the reasoning switch for one-line explanations behind each score.
* **Seamless Output:** Returns your original table + a `SCORE` column (and `REASON` column if enabled).

***

## 🏁 Getting Started

<Frame>
  <img src="https://mintcdn.com/nurturev/OitEIaKlfl7lCKCJ/images/AI%20Scorer%20node%20configuration%20screenshot.png?fit=max&auto=format&n=OitEIaKlfl7lCKCJ&q=85&s=23d67320d4df8e20e68195cbb1323a59" alt="AI Scorer node configuration screenshot" style={{ borderRadius: '0.5rem', width: '100%', margin: '1.5rem 0' }} width="1182" height="1664" data-path="images/AI Scorer node configuration screenshot.png" />
</Frame>

<Steps>
  <Step title="Add the Node">
    Drag & drop the **AI Scorer** node into your flow.
  </Step>

  <Step title="Select Inputs">
    Choose the columns to score and define your scoring criteria.
  </Step>

  <Step title="Configure Range">
    Set your **Min Score** and **Max Score** values.
  </Step>

  <Step title="Toggle Reasoning">
    (Optional) Turn on **Include Reasoning** to get short explanations.
  </Step>

  <Step title="Name Output">
    Rename the output column if needed (default: `SCORE`).
  </Step>

  <Step title="Run the Node">
    Hit **Run**—your table will now include `SCORE` (and `REASON` if enabled).
  </Step>
</Steps>

***

## Inputs

### 🛠️ Required Fields

* **Columns to Score (✅)**\
  Select which columns the AI should evaluate. Example: `region`, `engagement_score`, `plan_type`.

* **Scoring Criteria (✅)**\
  Write your rules in plain English. Example:\
  `"Score = max if Region = Americas AND Date is within last 30 days"`

* **Min Score (✅)**\
  The lowest value allowed (e.g., `0` or `1`).

* **Max Score (✅)**\
  The highest possible score (e.g., `100`).

### 🎯 Optional Fields

* **Include Reasoning (⚪️)**\
  Adds a `REASON` column that explains why each score was assigned.

* **Output Column Name (⚪️)**\
  Rename the `SCORE` column to fit your use case (e.g., `FitScore`, `RiskScore`).

***

## Output

You’ll get your original table with:

* `SCORE` (numerical output)
* `REASON` (optional one-liner explanation)

***

## How It Works

1. Extracts the selected fields for each row
2. Combines them with your natural-language criteria
3. Sends each prompt to the AI engine
4. Receives a numeric score (and reasoning if toggled)
5. Appends new output columns to your dataset

***

## 🚀 Use Cases & Prompts

| Use Case                    | Prompt Example                                                     |
| --------------------------- | ------------------------------------------------------------------ |
| Prioritize Enterprise Leads | “Score = max if PlanType = 'Enterprise' AND EngagementScore > 75.” |
| Churn Risk Scoring          | “Score = HealthScore × 0.6 − ChurnRiskScore × 0.4.”                |
| Flag At-Risk Accounts       | “Score = max if IsActive = false AND DaysSinceLastLogin > 30.”     |
| MRR-Based Fit Score         | “Score by descending MonthlyRecurringRevenue.”                     |
| Urgency-Based Routing       | “Score = max if RenewalDate is within next 14 days.”               |

***

## ✨ Pro Tips

<Tip>
  Use **2–3 key columns** for faster, more focused scoring.
</Tip>

<Tip>
  Toggle **Include Reasoning** to generate helpful notes for sales, CX, or ops to review.
</Tip>

<Tip>
  **Start small:** Test scoring logic on a sample table before deploying at scale.
</Tip>

***

## ⚠️ Important Considerations

<Warning>
  Every row processed consumes **1 credit**.
</Warning>

<Warning>
  Poor or null data can reduce scoring accuracy. Clean your input before running.
</Warning>

<Warning>
  Overly vague instructions may lead to default or uniform scoring. Be specific with thresholds and logic.
</Warning>

***

## 🛠 Troubleshooting & Gotchas

| Symptom                 | What’s Going On                  | Quick Fix                                |
| ----------------------- | -------------------------------- | ---------------------------------------- |
| All scores = Min or Max | Criteria too strict or too broad | Refine your logic to use tighter filters |
| Empty `SCORE` values    | Nulls in required columns        | Pre-clean data or add defaults upstream  |
| REASON column missing   | Reasoning toggle not enabled     | Turn on **Include Reasoning** in config  |

***

## 📝 FAQ

<AccordionGroup>
  <Accordion title="Can I chain multiple AI Scorer nodes?">
    Yes — for example, score for Fit first, then pass that into a Risk or Value scorer.
  </Accordion>

  <Accordion title="Can I reuse this logic later?">
    You can duplicate the node or save the flow as a template for later reuse.
  </Accordion>

  <Accordion title="Does this work on categorical and numeric fields?">
    Absolutely — the AI can reason across both types.
  </Accordion>
</AccordionGroup>

***

## 💰 Pricing

| Action         | Credit Cost |
| -------------- | ----------- |
| Per row scored | 1 credit    |

<Note>
  Only rows successfully scored are charged.
</Note>

***

<p style={{ fontSize: '1rem', fontWeight: 'bold', marginTop: '1.5rem' }}>
  Embed smart scoring anywhere in your flow — and turn raw records into ranked, reasoned insights that move your GTM faster. 📊⚡️
</p>