Gazelle AI

Ask a business question. Get a modeled answer.

Gazelle AI is the conversational layer across the platform. It pairs generative AI with NorthLight’s predictive models — so the answer you get is grounded in your modeled data, not a guess.

Plain language
questions, no query writing
Grounded
in your modeled data
Across
pricing, promotion and forecasting

The problem

Insight is stuck behind dashboards and analyst queues.

Executives ask “what if?” and wait days for an answer. Dashboards show what happened, but not why — and general-purpose AI tools can sound confident without any model behind the answer.

  • Every new question becomes a ticket for the analytics team.

  • Dashboards show numbers, not the reasoning behind them.

  • Generic AI chat tools aren’t connected to your models or data.

How it works

Generative AI on top of predictive models.

NorthLight is unusual in pairing the two. The language model asks the questions of the predictive models — it doesn’t invent the numbers.

  1. 01

    You ask in plain language

    Type an unstructured business question — about a price change, a promotion or a forecast.

  2. 02

    Gazelle uses your modeled data

    The question is answered from your input and modeled data — the elasticities, promotion effects and forecasts the platform keeps current.

  3. 03

    The analysis runs

    Gazelle runs the relevant pricing, promotion or forecast analysis on the models.

  4. 04

    You get the answer — and the why

    Interpretation, insights and implications are written for you automatically, in plain language.

Gazelle AI · how an answer is made
You ask“What happens to margin if we raise the 16 oz price by 3%?”

Generative AI

Understands the question, writes the explanation

Predictive models

Your elasticities, promotion effects & forecasts

You get
The numberThe whyThe implications

Schematic — not a product screenshot.

What you can ask

Questions it is built for

Price scenarios

What happens to margin and volume if we change a price?

Competitive gaps

Which products sit outside our competitive price band?

Promotion reviews

Which events paid back last quarter, and which did not?

Forecast drivers

What is behind the change in next quarter’s forecast?

“What happens to margin if we raise the 16 oz price by 3%?”“Which SKUs are under-priced versus competitors in the Northeast?”“Summarize how our Q3 promotions performed.”

Why it’s different

What sets Gazelle AI apart.

1

Grounded answers

Answers come from your modeled data, not general knowledge.

2

The why, not just the what

Every answer comes with interpretation and implications.

3

No query language

Ask the way you would ask a colleague.

4

One layer across all three

The same assistant works across pricing, promotion and forecasting.

See it in action

See Gazelle AI on a category like yours.

We show the real product on real data in a 30-minute walkthrough — not a mock-up.

Connected to

Works on the models you already run.

Reads from

Pricing modelsPromotion modelsForecasting modelsYour modeled data

Answers in

Plain-language summariesChartsFollow-up actions

Gazelle AI FAQ

Questions we hear about this.

Something else? info@northlightanalytics.com

Is Gazelle AI a general chatbot?+

No. It is connected to your predictive models and modeled data. It is designed to answer business questions about pricing, promotion and forecasting.

Where do the numbers come from?+

From the predictive models the platform estimates and keeps current on your data. The generative AI layer interprets and explains them.

Do we need to learn a query language?+

No. Ask unstructured questions in plain language.

Can we see it working on real data?+

Yes — book a demo and we will walk through Gazelle AI on a category like yours.

See it on your own data.

A 30-minute walkthrough of Gazelle on a category like yours.