AI Web Developer / AI Application Development needed for scoping over an existing extension product

Our product is a consumer shopping extension that helps buyers on the checkout page across a large catalogue of merchant storefronts. The UI and the backend are mostly mapped out. The hard part now sits in the AI layer that has to parse, classify and match offers across checkout pages that all look slightly different. We want an AI web developer team that has actually shipped AI application development into production, not a notebook demo. For this round we want your view on: - Architecture for parsing thousands of unfamiliar checkout pages - Classifier design for checkout-page elements - what is an offer block, what is a price, what is a coupon field - Prompt pipeline structure - versioning, prompt-as-code, deterministic fallbacks - Evaluation framework you would put around a parser at scale - Cost control and model swap - a clean abstraction between the AI layer and the underlying model We are happy to start with a paid scoping session. Our team works off structured written deliverables, so this round is fully written, no call needed. Output: a written addendum (5-10 pages) the build team can pick up. Existing product context (architecture, current parser draft, merchant integrations, internal metrics) and any access to the repo will be shared only with the team we sign the contract with, after NDA. We will not share specifics before contract. To apply, share 2-3 AI projects your team has shipped to production, a short note on how you would phase this AI layer from parser to evaluation framework, and your fixed price for this scoping round.