Predict.Engage.Immerse.

Platform

Operate personalisation from one control plane

NeuronSearchLab unifies catalog ingestion, event capture, and model oversight so teams can curate tenant-specific experiences without wrangling bespoke infrastructure.

Realtime learning

Event ingestion lands within seconds and each request recomputes embeddings with the latest contextual filters.

Privacy-first

Tenant-scoped credentials, OAuth client controls, and role-aware teams keep access limited to the right people.

Unified signals

Catalog items, user activity, and business rules converge into a shared embeddings store with built-in usage analytics.

Experiments

Coordinate rollout guardrails and targeting rules from the console before promoting updates live.

All platform capabilities

[01/08] // System init

NeuronSearchLabRecommendation Engine

A developer-first platform for real-time, context-aware recommendations – without building your own ML stack.

ProductNeuronSearchLab

[02/08]Intro & scroll

Viewport

Act on Event

Track every interaction – views, clicks, completion – and feed them straight into NeuronSearchLab so your recommendations feel alive with each scroll.

Key places: Home feed, watch-next, related content

[03/08]

RequestsAccuracyA/B TestsLift

Analytics

Track every call to the engine – requests, responses, and downstream events – so you can measure recommendation accuracy, understand request volume, and see what’s working in each placement. Run A/B tests, compare models, and prove uplift with real numbers instead of gut feel.

[04/08]Interaction

Control the Logic

“Blend neural embeddings with business rules – boost live events, hide watched items, and keep full control over what appears in every placement.”

[05/08]

Vector

Embeddings

How it works

NeuronSearchLab turns users, items into embeddings so your app can surface “what feels right” – not just “what matches a filter.”

[06/08]Use cases

Infinite loop

“Drop NeuronSearchLab into any app that has users and items – and turn static carousels into living, learning recommendation loops.”

[07/08]Cards

Rotate through multiple recommendation strategies – “because you watched”, “trending now”, “similar to this item” – all powered by one engine.

Add subtle motion and highlighting as scores update in real-time from NeuronSearchLab.

[08/08]

Smart feedsIn a few lines of code

Plug NeuronSearchLab into your API layer, send items and events, and start serving personalised rails, feeds, and “watch next” carousels in minutes.

Book a demo

NeuronSearchLab team

Built for media, sports & modern apps

Integration

How it works

Easily ingest content, stream events, get recommendations, and improve continuously, all in four simple steps.

1
Connect your content
Simply point us to your existing content feed or metadata - via MRSS, API, or SDK - and we'll instantly understand your items.
2
Define & Stream Events
Set the key actions that matter (views, clicks, purchases) and send them through our API or SDK.
3
Get Recommendations
Call /recommend with optional context to instantly deliver personalised, logic-driven results.
4
Learn & Improve
Our SDK captures rich behavioural data to power advanced analytics and continuously sharpen recommendations in real time.

What you'll pay

Choose the right plan for your rollout

Start with the recommendation volume and team access you need today. As traffic grows, move into deeper analytics, experimentation, and dedicated support without changing your integration.

Starter

For launching a first recommendation experience with managed infrastructure included.

£0/month

Up to 10k monthly requests

£0.05 per additional request

Best for:

10k recommendation requests included
1 live recommendation setup
2 recommendation contexts and core business rules
2 team seats with usage reporting

Also included

Production access
5 tracked event types
1 active model endpoint
7 days of analytics history

Choose Starter

Basic

For teams adding more recommendation surfaces and clearer result insight.

£0/month

Up to 25k monthly requests

£0.05 per additional request

Best for:

25k recommendation requests included
2 live recommendation setups
5 recommendation contexts and 25 business rules
5 team seats and 30 days of analytics

Also included

Production and staging access
10 tracked event types
1,500 recommendation explanations
Custom event types

Choose Basic

Pro

For scaled products that need experimentation, automation, and dedicated models.

Features built for personalization operators

NeuronSearchLab unifies data ingestion, context design, and release governance so product, growth, and engineering teams can steer recommendations without wrangling infrastructure.

NewModel launch control

Control every model launch from one console

Trigger tenant-specific training runs, review version metadata, and bake contextual guardrails into every release without swapping tools.

See docs Open console

Unified signals

Bring catalog attributes, behavioural events, and guardrail logic into one signal layer with usage analytics baked in.

Tenant credential control

Issue, rotate, and revoke API keys or OAuth clients while keeping tenant boundaries tight-no infrastructure tickets required.

Responsive recommendationsNew

Stream events in seconds, recompute embeddings per request, and kick off retraining runs when behaviour shifts.

Stand up pipelines in days

Push catalog entities and events through ready-made APIs, validate payloads in the console, and track training progress as you ramp.

Coordinate rollouts

Prepare guardrails, targeting rules, and approvals before any model goes live, all within the same workspace.

What you'll see

Adaptive relevance in action

This sandbox simulates a personalised feed. A fake cursor hovers, clicks “Add to cart,” and favourites items, emitting lightweight and strong signals that re-rank the “For you” grid in real time. Toggle User X/Y to compare starting preferences and watch relevance update as events roll in.

home-feed

Personalised home

Trending