
The problem
Product discovery requires aligning open-ended intent handling with current, structured catalogue data.
A retail discovery assistant has to connect a shopper’s language to products the business actually carries. The useful engineering boundary is the path from intent to structured retrieval, synchronized catalogue state and a result the application can present consistently.
My responsibility
Scoped AI architecture, retrieval workflow and system integration—not ownership of Rahva Raamat’s wider commerce platform.
The engineering
Production AI infrastructure combining intent processing, retrieval workflows, structured presentation and catalogue synchronization to support catalogue-grounded discovery.
A catalogue-grounded discovery architecture that aligns unstructured query intent with structured product retrieval and catalogue state.
The delivered result
A catalogue-aware path from open-ended query to structured product results. The public storefront verifies the catalogue context; no conversion, revenue or performance claim is presented.
Sources & scope
Selected production infrastructure · public catalogue context. The official product link documents the business and platform context. The responsibility and implementation described here are my case study account.
