Semantic layerLive

One source of truth behind every number.

We define every number for you, so every team reads the same one. Built for India, with nothing to set up.

01Where one number comes from

Contribution margin arrives as four messy fields.

Four tools, four formats, four definitions of the same thing. The layer reconciles them once. After that, every dashboard and Ocular AI read one number.

Raw, as it lands
  • total_priceShopify

    tax + shipping baked in

  • referral_feeAmazon

    per ASIN, settled late

  • rtoShiprocket

    returns still in transit

  • settlementGoKwik

    COD and prepaid mixed

The layer reconciles
  • Every amount converted to ₹
  • Every timestamp moved to IST
  • Field names unified across tools
  • RTO and returns subtracted, not ignored
  • COD and prepaid split apart
One measure
contribution_margin

Revenue, net of fees, ad spend, shipping and returns. In ₹. By SKU, channel, pincode or occasion.

Read by, identically
  • P&L dashboard
  • Ocular AI
  • Every team
02

What the layer does

One definition across every channel you sell on

Meta, Google, Amazon, Flipkart, Blinkit and your own storefront name the same thing five different ways. The layer reconciles naming, currency and structure once, so a number means one thing everywhere.

MetaGoogleAmazonFlipkartBlinkitStorefrontone definition

Raw fields become margin-aware measures

Not cleaned-up columns. Measures that already net out fees, returns and shipping the way a P&L would.

The logic is already built

Hundreds of measures and dimensions come modelled and ready, with the rules, fees, returns and currency worked out for India. You query and slice them. You never have to build the logic.

Every team reads the same number, no SQL

Finance, growth and ops query trusted measures in plain language. Nobody re-derives contribution margin in a spreadsheet, because there is only one to read.

Finance
Growth
Ops
03

Why this matters for Ocular AI

Try this question

"Did Tier-2 returns eat the margin from our Diwali Meta campaigns, and which Nykaa POs are at risk this week given Blinkit availability in Mumbai?"

Every concept in that sentence already exists in the model. Tier-2. Returns. Diwali. Meta-attributed margin. Nykaa PO. Dark-store availability. Ocular AI reads the model. It answers.
See it in Ocular AI

Read the model. See the moat.

Thirty minutes on your data, or ours. We walk the model end-to-end, then you ask any question you have got.