Skip to main content

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.

These walkthroughs show how the ranking platform features work together to solve real problems. Each scenario can be completed entirely within the console — no engineering support required.

​
Scenario 1: newsroom weather emergency

Goal: Pin a closures guide to slot 2 on the homepage during a live weather event.
1

Create the rule

Open Console > Rules Engine, click New Rule, and create a pin rule named Pin school closures guide to slot 2.
  • Condition field: entity_id
  • Operator: equals
  • Value: itm_school_closures_live
  • Action: pin the item to position 2
  • Scope: homepage context
  • Status: active
2

Verify with Explainability

Open Console > Explainability, enter a reader ID and the item ID itm_school_closures_live, then click Explain.Pass condition: Explainability shows the rule Pin school closures guide to slot 2 matched the item.
3

Confirm the pipeline

Open Console > Pipeline Config and confirm the active pipeline for the homepage context has the Rules stage enabled. Also confirm no experiment is excluding the rule.
This scenario is a good operational test because only one valid result exists. If the item is not in slot 2, the change failed.

​
Scenario 2: Streaming platform series launch

Goal: Promote a new series for two weeks, targeting users who haven’t watched it yet, while running an A/B test to measure the impact.
1

Define the audience

Create two segments in Console > User Segments to separate users who have and have not engaged with the series.Segment: “Hasn’t watched Series X”
  • Type: item_interaction
  • Item ID: the series item ID, for example itm_series_x_ep1
  • Operator: interactions <
  • Value: 1
  • Name: “Hasn’t watched Series X”
Segment: “Started but didn’t finish Series X”
  • Condition 1: Type item_interaction, Item ID series_x_ep1, Operator interactions >, Value 0
  • Condition 2: Type item_interaction, Item ID series_x_ep10, Operator interactions <, Value 1
  • Logic: AND
  • Name: “Started not finished Series X”
2

Create the rules

Create two scheduled rules in Console > Rules Engine.Rule: “Pin Series X for new viewers”
  • Type: pin
  • Priority: 80
  • Condition: segment_id equals “Hasn’t watched Series X”
  • Action: pin the series into the top 3 positions
  • Schedule: campaign launch through two weeks later
Rule: “Boost continuation for partial viewers”
  • Type: boost
  • Priority: 70
  • Condition: segment_id equals “Started not finished Series X”
  • Action: boost the series episodes with a high factor
  • Schedule: same campaign window
3

Run an A/B test

Go to Console > A/B Testing, click New Experiment, and create “Series X launch campaign”.
  • Description: “Hypothesis: pinning Series X for non-viewers increases completion rate.”
  • Control (50%): no config overrides; standard recommendations
  • Treatment (50%): config overrides with include_rule_ids set to the two rule IDs above
  • Status: Running on launch day
4

Measure results

During the campaign, open the experiment’s Results tab and click Refresh metrics periodically. Compare CTR and conversion rate between Control and Treatment, then check lift percentage.After two weeks, set the experiment status to Completed. The rules auto-deactivate via their schedule.

​
Scenario 3: E-commerce flash sale weekend

Goal: Boost sale items for a weekend, show premium items to high-value customers, suppress out-of-stock products, and test whether manual merchandising rules outperform pure ML.
1

Create the high-value customer segment

Go to Console > User Segments, click New Segment, and create “High-value customers”.
  • Type: computed
  • Field: total_events
  • Operator: greater_than
  • Value: 100
2

Create the sale rules

Create the merchandising rules in Console > Rules Engine.Rule: “Flash sale boost”
  • Type: boost
  • Priority: 60
  • Condition: category equals sale
  • Action: boost with a high factor
  • Schedule: Friday 18:00 to Sunday 23:59
Rule: “VIP exclusive items”
  • Type: boost
  • Priority: 80
  • Conditions: segment_id equals “High-value customers” and tier equals premium
  • Action: boost premium items strongly
  • Schedule: same weekend window
Rule: “Suppress out-of-stock”
  • Type: filter
  • Priority: 100
  • Condition: stock_status equals out_of_stock
  • Action: exclude matching items
  • Schedule: none; always active
3

Set up the pure ML experiment

Go to Console > Pipeline Config, create or note your default pipeline, and consider creating a second pipeline with the rules stage disabled for pure ML ranking.Then create an experiment in Console > A/B Testing:
  • Control (50%): default pipeline
  • Treatment (50%): pipeline with rules stage disabled
  • Status: Running on Friday
4

Monitor the sale

Open Console > Analytics and watch served volume. Filter by user ID to spot-check that VIP users see premium items, and refresh experiment metrics throughout the weekend.After the sale, complete the experiment and compare conversion rates. The sale rules auto-deactivate after Sunday, so no cleanup is needed.

​
Scenario 4: Content freshness and diversity

Goal: Ensure recommendations always include recent content and don’t over-represent a single category.
1

Create freshness and diversity rules

Create both rules in Console > Rules Engine.Rule: “Boost new content”
  • Type: boost
  • Priority: 50
  • Condition: published_days_ago is less than 7
  • Action: boost by a moderate factor
  • Schedule: none; always active
Rule: “Diversify by category”
  • Type: diversify
  • Priority: 40
  • Conditions: none; applies to all results
  • Action: limit to maximum 3 items per category value
  • Schedule: none; always active
2

Verify with Explainability

Go to Console > Explainability, enter a test user ID and an item ID for a recently published item, and confirm the “Boost new content” rule shows matched.Try an older item and confirm it shows no match.

​
Scenario 5: Gradual feature rollout

Goal: Roll out a new set of ranking rules to 10% of users first, then expand.
1

Deploy rules as inactive

Create your new rules but leave them inactive with the toggle off. Note their rule IDs.
2

Create a staged experiment

Go to Console > A/B Testing and create an experiment:
  • Control (90%): no config overrides
  • Treatment (10%): config overrides with include_rule_ids set to the new rule IDs
  • Status: Running
3

Monitor and expand

Check experiment metrics after a few days. If metrics look good, edit the experiment and adjust traffic to Control 50% and Treatment 50%.Continue monitoring. When confident, set the experiment to Completed, activate the rules for everyone, and delete the experiment.

​
Combining features

These scenarios demonstrate a pattern: segments define who, rules define what, scheduling defines when, pipelines define how, and experiments measure whether it works.
FeatureRole
User SegmentsTarget specific user cohorts
Rules EngineOverride rankings with business logic
Rule schedulingTime-bound campaigns
Pipeline ConfigControl which processing stages run
A/B TestingMeasure impact with traffic splits
ExplainabilityDebug and verify before launch
AnalyticsMonitor outcomes in production