A WhatsApp AI agent for e-commerce sells from a live catalog: it finds the variant, confirms stock, quotes shipping, creates the order and hands off to a person when needed.
KINTO’s cases show how this works in practice and where it stops.
A store with a deep catalog does not have the same problem as a company that answers a handful of repeated questions. The customer asks for a specific model in a specific size, or a part for a specific vehicle and year. A wrong answer is not a small embarrassment: it is a returned product, an angry customer or a sold-out item sold twice.
The page therefore answers a narrower question than the general WhatsApp agent page: can an agent handle a catalog with variants and live stock, quote delivery and create the order in the store?
Sports e-commerce. A sporting goods brand had complex variants and sold mostly over WhatsApp. Per the owner, order automation was very good but chat was the weak point: slow replies, wrong sizes, out-of-stock items sold and orders that never reached the store. KINTO built an agent with live stock and a curated spec sheet. Details are in the case study. It converted about 9 to 10% of conversations in its first weeks, and the owner reported channel sales about 30% above history in the first full month.
Auto parts. A distributor with over 10,000 products online had inventory in an ERP that the store never saw. KINTO made the ERP the source of truth for the agent. The case study reports about 26% more sellable products and hundreds of phantom products no longer offered. The agent searches by vehicle, part or part number, quotes shipping by zone, pre-registers branch pickups and alerts the team.
For the sports brand, the ad-spend savings in absolute terms and the agent’s effect separated from seasonality are not known, and unifying product codes between the ERP and the store is still open. For auto parts, there is not enough evidence of impact on sales, and catalog images remain the main debt.
The general method is on the AI agents for WhatsApp page. For the store side, see headless e-commerce.
WhatsApp & voice agents
KINTO fixed the weak link of an e-commerce brand that sells over WhatsApp with a multi-model audited agent and order-by-order reconciliation; the owner reported a month with channel sales ≈30% above their history.
Headless e-commerce
By making the ERP, not the web catalog, the source of truth, KINTO got an auto parts distributor’s WhatsApp agent to offer ≈26% more real products and stop offering hundreds it did not have.
It looks each variant up instead of remembering it. For a sporting goods brand with model, size, color and stock combinations, the agent works from live stock and a curated spec sheet. If the sheet has no data, the agent says so, and orders reach the store instead of getting lost in the chat.
By checking compatibility against model-year data and by looking up availability. In the auto parts case, the agent searches by vehicle, part or part number, confirms availability against ERP data synced daily and rejected a part whose year did not match the vehicle. It also recognized an image that was not a spare part.
In the auto parts case it quotes shipping by zone and pre-registers branch pickups. It also checks payments without being able to change them, and alerts the team, so money decisions stay with a person while the quote and the order move without waiting for an office.
Modest, measured numbers. The sports brand's agent closed about two dozen orders at roughly 10% conversion in its first weeks, and the owner reported channel sales about 30% above history. The auto parts agent had few, small orders in its first weeks, so its effect on sales is not yet shown.
Tell KINTO what is going on. The first conversation is to understand it, not to sell.
