---
title: "Create recommender"
method: POST
path: "/recommenders"
tags: ["Recommenders"]
---

# Create recommender

`POST /recommenders`

Create a new Spark recommender for a specific shop and product feed. Training starts automatically after creation.

## Request body

- object
  - `shop_id` string, required — The shop (e-commerce platform) identifier. Max 64 characters, allowed characters: letters, digits, `.`, `_`, `-`.
  - `feed_id` integer, required — The product feed ID to use for recommendations.
  - `name` string — Optional human-readable name of the recommender. Max 191 characters.
  - `filtered_products` string — Optional list of product identifiers (SKUs) to exclude from the recommendations. Max 1000 items, each unique and max 255 characters. Omit the field for no product filter.
  - `filtered_categories` string — Optional list of category names to exclude from the recommendations. Max 1000 items, each unique and max 255 characters. Omit the field for no category filter.
  - `included_months` string — Optional list of calendar month numbers (`1`–`12`) that restrict which purchase history is used for training. Values must be unique, max 12 items. Omit the field to use the full purchase history.

## Response `200`

OK

- object[]
  - `recommender_id` string
  - `feed_id` integer
  - `shop_id` string
  - `last_trained_timestamp` unknown
  - `status` string
  - `settings` object
    - `filtered_products` string[]
    - `filtered_categories` string[]
    - `included_months` integer[]
  - `sequence_number` integer
  - `name` string

## Other responses

- `422` — Unprocessable Entity

---

[API](https://skmtc.dev/ecomail/apis/ecomail-api.md) · [All operations](https://skmtc.dev/ecomail/apis/ecomail-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/ecomail/ecomail-api/revisions/31b5789931fe/schema)