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

# Recommendation Control Demo Path

> A buyer-facing walkthrough that proves NSL can control, explain, and improve a recommendation surface.

Use this path when you want to show that NeuronSearchLab is more than a recommendation API. The demo should prove that a team can define a surface, tune ranking behavior, apply business logic, and explain the result.

The strongest version is a vertical-specific walkthrough, such as news/media, commerce, sport, or video. Keep the story narrow: one surface, one goal, one ranking change, one explanation.

## Demo story

<Steps>
  <Step title="Start with a surface">
    Open or create a context for a recognizable surface, such as `Homepage Feed`, `Related Articles`, `Product Detail Rail`, or `Video Continue Watching`.
  </Step>

  <Step title="State the business goal">
    Explain what the surface is meant to improve: discovery, CTR, conversion, retention, long-tail coverage, freshness, or editorial control.
  </Step>

  <Step title="Show the recipe">
    Open the context's Recipe tab and show how candidate sources, signals, discovery settings, and guardrails are attached to that surface.
  </Step>

  <Step title="Show the pipeline">
    Open the Pipeline tab and show the runtime stages: candidate generation, enrichment, scoring, rules, ranking, and post-processing.
  </Step>

  <Step title="Apply one business control">
    Use a simple rule, such as boosting fresh content, filtering unavailable items, capping sponsored items, or pinning a launch item.
  </Step>

  <Step title="Explain the result">
    Use Explainability to show retrieval score, rule impact, pipeline stage status, and feature contributions for one user-item pair.
  </Step>
</Steps>

## What the buyer should understand

By the end, the buyer should believe:

* NSL can separate ranking behavior by product surface.
* Operators can adjust ranking without waiting for application releases.
* Business rules sit after model scoring, so control does not replace relevance.
* Pipelines expose the runtime path instead of hiding everything in a black box.
* Explainability makes recommendations inspectable enough for debugging, demos, and rollout confidence.

## Suggested proof scenario

For a media/news demo:

1. Create or open a `Homepage Feed` context.
2. Attach a discovery-oriented recipe.
3. Show candidate generation and rules stages in the pipeline.
4. Add a rule that boosts recent editorial priority content but keeps relevance in the ranking.
5. Request recommendations for a sample user.
6. Explain one recommended item and point to the score, matched rules, and pipeline stages.

The message is simple: NSL gives teams inspectable control over recommendation behavior, not just a black-box ranked list.

## Related pages

* [Contexts](/contexts)
* [Console Contexts](/guides/contexts)
* [Pipeline Config](/guides/pipeline-config)
* [Rules Engine](/guides/rules-engine)
* [Explainability](/guides/explainability)