Context
Multiple brand storefronts, each with its own catalog and customer base, sharing one search implementation. The work covers core product search, an AI-assisted search experience, and Algolia-powered recommendations.
The AI search drawer, with guardrails
The AI search drawer gives shoppers a conversational search surface. The part I would call out is the guardrail classification layer in front of it: a classifier decides which queries the AI surface should answer and which should fall through to standard results. On catalogs that include weapon-adjacent products, an unguarded AI answer surface is a liability, so the guardrail layer is not a nice-to-have. It is the thing that made the feature shippable. [SIGN-OFF: confirm you are comfortable describing the guardrail layer publicly at this level of detail.]
Cross-sell carousel
An Algolia-powered cross-sell carousel surfaces related products on [CONFIRM: where it renders, e.g. product pages, the cart drawer, or both].
Outcome
[METRIC: search conversion rate, null-result rate reduction, cross-sell attach rate, or AI drawer engagement, whichever you can actually source. Delete any you cannot back up.]