# Tamr developer docs > Tamr's developer docs: the restaurant ordering and AI waiter SDK for the web, React Native, Flutter, iOS and Android, the HTTP API (/api/v1) and webhooks. Diners chat with the restaurant's AI waiter inside the restaurant's own app; the app's backend signs diner tokens, sends orders and receives signed webhooks. ## Get started - [Overview](https://tamr.link/docs.md): Put Tamr's AI waiter in your restaurant app with the restaurant ordering SDK for the web, React Native, Flutter, iOS and Android, plus the API reference and webhooks. - [How it works](https://tamr.link/docs/how-it-works.md): How Tamr's AI waiter SDK works: a hosted waiter page in a webview, a typed bridge to your app, diner identity from your backend, and the Client and Server APIs. - [Keys and test mode](https://tamr.link/docs/keys.md): Publishable app keys and secret API keys for Tamr's SDK: test keys versus live keys, guest mode versus signed-in diners, and what the restaurant sets up for you. - [Quickstart](https://tamr.link/docs/quickstart.md): Add Tamr's AI waiter to a web, React Native, Flutter, iOS or Android app in five minutes: get a test key, install the SDK, open the waiter, then sign diners in. ## SDKs - [Overview](https://tamr.link/docs/sdks.md): Tamr's AI waiter SDKs for the web, React Native and Expo, Flutter, iOS (Swift) and Android (Kotlin): packages, platform floors and the shape they share. - [Web](https://tamr.link/docs/sdks/web.md): Add Tamr's AI waiter to a restaurant website with @tamr/waiter-web or one script tag: a launcher, a side sheet or inline, signed-in diners, cart lines and events. - [React Native](https://tamr.link/docs/sdks/react-native.md): Add Tamr's AI waiter to a React Native or Expo app with @tamr/waiter-react-native: install, the waiter host, signed-in diners, cart lines, voice notes and debug mode. - [Flutter](https://tamr.link/docs/sdks/flutter.md): Add Tamr's AI waiter to a Flutter app on iOS and Android with tamr_waiter: install, open the waiter as a sheet, page or widget, sign diners in and take cart lines. - [iOS](https://tamr.link/docs/sdks/ios.md): Add Tamr's AI waiter to an iOS app in Swift with TamrWaiter: Swift Package Manager, UIKit and SwiftUI, signed-in diners, cart lines, voice notes and debug mode. - [Android](https://tamr.link/docs/sdks/android.md): Add Tamr's AI waiter to an Android app in Kotlin with link.tamr:waiter-android: Maven Central, Compose and Views, signed-in diners, cart lines, voice notes and debug mode. - [Events](https://tamr.link/docs/sdks/events.md): The events Tamr's AI waiter SDK sends your app: ready, state, order.placed, order.cancelled, payment.completed, checkout, open_dish, bookings, staff calls and logs. - [Look, wording and language](https://tamr.link/docs/sdks/appearance.md): Match Tamr's AI waiter to your app at runtime: colours, fonts, corners, the waiter's name and labels, starters, diet and allergen restrictions, theme, language and context. ## Guides - [Overview](https://tamr.link/docs/guides.md): Guides for Tamr's restaurant ordering SDK: diner identity, ordering modes, test orders, product ids and nutrition, treats, reviews, custom chats, debugging and going live. - [Diner identity](https://tamr.link/docs/guides/diner-identity.md): Sign a short EdDSA JWT on your backend to tell Tamr's AI waiter which of your users is chatting, in Node, Kotlin, Dart or Swift, or mint a one-time launch code instead. - [Ordering modes](https://tamr.link/docs/guides/ordering.md): How ordering works with Tamr's AI waiter SDK: dine-in and pickup orders through Tamr with Tamr Pay, or the waiter's dishes proposed into your app's own cart and checkout. - [Sending your orders](https://tamr.link/docs/guides/sending-orders.md): When orders go to your app's cart, send each order your checkout placed to Tamr from your backend, so the AI waiter can offer the diner's usual and one-tap reorder. - [Test orders](https://tamr.link/docs/guides/test-orders.md): Try a pickup order end to end with a Tamr test key: priced and checked like a real order, sent to your test webhook endpoints, never to the restaurant's kitchen. - [Product ids and nutrition](https://tamr.link/docs/guides/product-ids-and-nutrition.md): Map your own SKUs to the restaurant's dishes, sizes and options so cart lines and webhooks name your products, and send nutrition figures for the dishes Tamr has none for. - [Keep in step with the menu](https://tamr.link/docs/guides/keeping-in-step.md): Make Tamr the source of truth for your menu and prices: read each menu from your backend, store its version, refresh on the menu.changed webhook and poll as a fallback. - [Treats on the house](https://tamr.link/docs/guides/treats.md): Handle a treat on the house from Tamr's AI waiter in your own checkout: price the treat's cart line at zero, then redeem it from your backend once the order is placed. - [Post-meal reviews](https://tamr.link/docs/guides/reviews.md): Ask diners to rate each dish of an order your backend sent, 1 to 5 stars, from your app's AI waiter chat, and receive the ratings as review.submitted webhooks. - [Hiding features](https://tamr.link/docs/guides/hiding-features.md): Turn off Tamr's post-meal review ask, take-home suggestions, treats on the house or voice notes for an app that has its own, from the dashboard or from your app. - [Build your own chat](https://tamr.link/docs/guides/custom-chat.md): Draw a native AI waiter chat on Tamr's Client API: start a session, stream the waiter's replies as server-sent events, draw dish cards, and renew expired sessions. - [Debug mode](https://tamr.link/docs/guides/debugging.md): Turn on debug mode in Tamr's AI waiter SDK to log every bridge message, request and tap between the waiter page and your app, and inspect the webview in debug builds. - [Erasing a user](https://tamr.link/docs/guides/erasing-a-user.md): Delete everything Tamr holds for one of your users with one API call, and how long Tamr keeps conversations, orders and webhook deliveries. - [Going live](https://tamr.link/docs/guides/going-live.md): The checklist before you release Tamr's AI waiter in your app: a live key, signing keys, webhooks, ordering, erasure, debug mode off and one real order end to end. ## API reference - [Overview](https://tamr.link/docs/api.md): The Tamr API reference: every /api/v1 route for your backend and for apps that draw their own AI waiter chat, with authentication, schemas and curl, Node and Python examples. - [Create a launch code](https://tamr.link/docs/api/create-launch-code.md): Mint a one-time launch code for one of your users from your backend, for apps whose backend can't sign diner tokens. - [Erase a user](https://tamr.link/docs/api/erase-user.md): Delete everything Tamr holds for one of your users: profile, sessions, conversations, ratings, orders and webhook deliveries. - [Send an order](https://tamr.link/docs/api/send-order.md): Send an order your own checkout placed to Tamr, so the AI waiter can offer the diner's usual, reorders and post-meal reviews. - [Redeem a treat](https://tamr.link/docs/api/redeem-treat.md): Redeem a treat on the house once your checkout has placed the order that priced it at zero. - [Set your product ids](https://tamr.link/docs/api/set-external-ids.md): Map your own SKUs and product ids to the restaurant's dishes, sizes and options, up to 500 at a time. - [Set nutrition figures](https://tamr.link/docs/api/set-nutrition.md): Send your own calories, protein, carbs and fat per dish to fill the figures the restaurant left empty in Tamr. - [Read a menu from your backend](https://tamr.link/docs/api/get-catalog-menu.md): Read one of the restaurant's published menus from your server, with prices, sizes, choices, availability, your product ids and the menu's version, to keep your own catalog in step with Tamr. - [Start a session](https://tamr.link/docs/api/start-session.md): Exchange a diner token, a launch code or a guest's device id for the session every other Client API call carries. - [End a session](https://tamr.link/docs/api/end-session.md): End the diner's session when they log out of your app, so the next sign-in starts a new conversation. - [Get the waiter](https://tamr.link/docs/api/get-waiter.md): Read the waiter's public face: name, greeting, starters, languages, brand, the restaurant's published app settings and where orders go. - [Get the menu](https://tamr.link/docs/api/get-menu.md): Read the menu the AI waiter serves at a branch and service mode, with Tamr's ids for every dish, size and option. - [Warm up the waiter](https://tamr.link/docs/api/warm-up.md): Prime the AI waiter before the diner's first message so the first answer comes faster. - [Send a message](https://tamr.link/docs/api/chat.md): Send the diner's message to the AI waiter and stream the reply as server-sent events: text, dish cards and the turn's end. - [List conversations](https://tamr.link/docs/api/list-conversations.md): List the diner's recent conversations with the AI waiter, newest first. - [Get a conversation](https://tamr.link/docs/api/get-conversation.md): Read one of the diner's conversations with every message and card, to redraw a chat after a reload or on another device. - [Close a conversation](https://tamr.link/docs/api/close-conversation.md): Close a conversation so the diner's next message starts a new one. - [Transcribe a voice note](https://tamr.link/docs/api/transcribe.md): Turn the diner's voice note into text, with the menu's dish names as hints, before sending it as a message. - [Match your cart](https://tamr.link/docs/api/match-cart.md): Match your app's cart lines to the dishes the AI waiter serves, by Tamr's ids or your own; unplaceable lines come back unmapped. - [Get the diner's usual](https://tamr.link/docs/api/get-usual.md): Read the signed-in diner's usual order, or their last one, from the orders your backend sent. - [Get the nutrition goal](https://tamr.link/docs/api/get-goal.md): Read the nutrition goal Tamr keeps for the signed-in diner: cutting, maintain or bulking, per meal. - [Set the nutrition goal](https://tamr.link/docs/api/set-goal.md): Set or clear the diner's nutrition goal per meal; the AI waiter checks dishes against it using the restaurant's own figures. - [Match an order for review](https://tamr.link/docs/api/match-review.md): List the dishes of an order your backend sent that the diner can rate, with any stars they already gave. - [Rate an order's dishes](https://tamr.link/docs/api/submit-reviews.md): Send the diner's 1 to 5 star ratings of an order's dishes, one review per dish per order. - [Seat the diner at a table](https://tamr.link/docs/api/seat-at-table.md): Seat the diner at the table whose QR code they scanned, checked against the restaurant's floor plan, before they order. - [Get the table](https://tamr.link/docs/api/get-table.md): Read whether the seated diner may send an order now, and the table's rounds so far. - [Send a table order](https://tamr.link/docs/api/send-table-order.md): Send the seated diner's order to the restaurant's team, with sizes, choices and notes, when orders go to Tamr. - [Get pickup options](https://tamr.link/docs/api/get-pickup-options.md): Read where and when the diner can collect a pickup order: locations, times and how each takes payment. - [Place a pickup order](https://tamr.link/docs/api/place-pickup.md): Place the diner's pickup order for collection at a branch, paid at the counter or ahead with Tamr Pay. - [List pickup orders](https://tamr.link/docs/api/list-pickups.md): List the diner's pickup orders, newest first, with their status and collection code. - [Get payment options](https://tamr.link/docs/api/get-pay-options.md): Read whether the diner can pay with Tamr Pay here, the providers, limits and tip presets. - [Start a payment](https://tamr.link/docs/api/start-payment.md): Start paying the table's tab, a typed amount or a pickup with Tamr Pay, and get the provider's payment page. - [Get a payment](https://tamr.link/docs/api/get-payment.md): Read where one of the diner's Tamr Pay payments stands: open, paid, failed, expired or refunded. - [Schemas](https://tamr.link/docs/api/schemas.md): The named schemas of Tamr's API: chat requests and events, dish cards, the menu, conversations, errors and every webhook payload, field by field. ## Webhooks - [Overview](https://tamr.link/docs/webhooks.md): Tamr webhooks tell your backend what happened in your app's AI waiter chats and on the restaurant's menus: orders placed and cancelled, payments, bookings, staff calls, reviews, chats ending and menu changes, signed per Standard Webhooks. - [Verify signatures](https://tamr.link/docs/webhooks/verify-signatures.md): Check every Tamr webhook's HMAC-SHA256 signature on the raw body before trusting it, with a Standard Webhooks library or a few lines of Node. - [Retries](https://tamr.link/docs/webhooks/retries.md): How Tamr retries a failed webhook delivery over three days, what turns an endpoint off, and how to de-duplicate and order events. - [order.placed](https://tamr.link/docs/webhooks/order-placed.md): The order.placed webhook: a diner's dine-in or pickup order from your app reached the restaurant's team, with ids, lines and amounts. - [order.cancelled](https://tamr.link/docs/webhooks/order-cancelled.md): The order.cancelled webhook: an order your app heard placed was declined, cancelled or expired before it was served. Undo it in your records. - [payment.completed](https://tamr.link/docs/webhooks/payment-completed.md): The payment.completed webhook: a diner paid with Tamr Pay from your app, recorded from the payment provider's own answer. - [cart.proposed](https://tamr.link/docs/webhooks/cart-proposed.md): The cart.proposed webhook: the AI waiter proposed lines to your app's cart, with Tamr's ids and yours, quantities and prices. - [staff.requested](https://tamr.link/docs/webhooks/staff-requested.md): The staff.requested webhook: the AI waiter called restaurant staff, for a routine request such as water or the check, or an escalation. - [booking.made](https://tamr.link/docs/webhooks/booking-made.md): The booking.made webhook: the AI waiter booked a table in your app's chat, with the branch, date, time and party size. - [booking.changed](https://tamr.link/docs/webhooks/booking-changed.md): The booking.changed webhook: the AI waiter moved or cancelled a table booking in your app's chat. - [review.submitted](https://tamr.link/docs/webhooks/review-submitted.md): The review.submitted webhook: a signed-in diner rated a dish of an order your backend sent, 1 to 5 stars. - [conversation.ended](https://tamr.link/docs/webhooks/conversation-ended.md): The conversation.ended webhook: a chat with the AI waiter in your app ended, resolved, idle, restarted or logged out. - [menu.changed](https://tamr.link/docs/webhooks/menu-changed.md): The menu.changed webhook: one of the restaurant's menus, its prices, its dishes or your product ids changed. Read the menu again from your backend. ## Resources - [Errors and rate limits](https://tamr.link/docs/errors.md): Every error code Tamr's API and SDK answer, what each means and what to do, plus the rate limits for live keys, test keys, guests and secret API keys. - [Changelog](https://tamr.link/docs/changelog.md): What changed in each release of Tamr's SDKs for the web, React Native, Flutter, iOS and Android, and of @tamr/api, from each package's own CHANGELOG. - [Support policy](https://tamr.link/docs/support-policy.md): How Tamr versions its SDKs and API: semantic versioning, the oldest platforms each SDK supports, how /api/v1 grows without breaking, and how to get help. ## More - [Every page in one file](https://tamr.link/docs/llms-full.txt) - [The OpenAPI spec](https://tamr.link/api/v1/openapi.json)