· 1 observation
Technical updates · 1
Package or technical metadata updated
View technical comparison →Eqolux v0.1.2
Analyze your purchasing data
From the developer
Not specified in the available declarations: capabilities, limitations, pricing & account requirements.
Who were my 10 biggest suppliers over the last 12 months?
Developer declarations; not independently tested. Listing last checked Nov 21, 2026. Package observed Nov 21, 2026.
From the marketplace listing
Eqolux is a purchasing analytics platform for restaurant and hotel groups: it reads their supplier invoices, delivery notes and credit notes and turns them into structured purchase data. This plugin connects your Eqolux account so you can analyze that purchasing data in plain language. Ask who your biggest suppliers are, how spend evolves by establishment and month, which product prices went up, where the same product is bought at very different prices, which purchases exceed a negotiated price, where your products come from and what their carbon footprint is. Answers link back to the products, documents and suppliers in the Eqolux app. The plugin adds two skills, purchasing analysis and savings opportunities, on top of the Eqolux connector. It only reads data: every answer is limited to the organizations and establishments your Eqolux account can access. Requirements: an Eqolux account. Sign in with your Eqolux credentials when prompted.
Language: English · Automatically detected from descriptions.
Exact text from the indicated source. A mention alone does not establish support for your task.
Analyze your purchasing data
Eqolux is a purchasing analytics platform for restaurant and hotel groups: it reads their supplier invoices, delivery notes and credit notes and turns them into structured purchase data. This plugin connects your Eqolux account so you can analyze that purchasing data in plain language. Ask who your biggest suppliers are, how spend evolves by establishment and month, which product prices went up, where the same product is bought at very different prices, which purchases exceed a negotiated price, where your products come from and what their carbon footprint is. Answers link back to the products, documents and suppliers in the Eqolux app. The plugin adds two skills, purchasing analysis and savings opportunities, on top of the Eqolux connector. It only reads data: every answer is limited to the organizations and establishments your Eqolux account can access. Requirements: an Eqolux account. Sign in with your Eqolux credentials when prompted.
· 1 observation
Package or technical metadata updated
View technical comparison →Finding similar plugins… Results will appear here automatically.
Browse plugins →---
name: eqolux-purchasing-analysis
description: Analyze restaurant and hotel purchasing data in Eqolux — spend by supplier, establishment, category or period, product prices and their evolution, origins and carbon footprint. Use whenever the user asks about their purchases, suppliers, invoices, delivery notes, products, prices, food cost or CO2 tracked in Eqolux.
license: MIT
metadata:
version: '0.1.0'
author: Eqolux
---
# Eqolux purchasing analysis
Eqolux reads supplier documents (invoices, delivery notes, credit notes) for restaurant and hotel groups and turns them into structured purchase data. The Eqolux connector provides three tools:
- `list_organizations`: the organizations (groups) the user belongs to, with their ids
- `describe_data`: the tables `run_sql_query` can read, every column, and how amounts are converted and signed
- `run_sql_query`: one read-only PostgreSQL `SELECT` on one organization's data, at most 500 rows
## Workflow
1. **Pick the organization.** Call `list_organizations`. If there is one, use it. If there are several, use the one the user named (or the one their establishment belongs to); otherwise ask. Each query covers exactly one organization.
2. **Read the schema once.** Call `describe_data` before the first query of the conversation.
3. **Query.** Aggregate in SQL rather than fetching rows to aggregate yourself: results stop at 500 rows. One well-built query beats several small ones. If a query times out, narrow the period or filter on a supplier, category or establishment.
4. **Answer** following the rules below.
## Rules that keep the numbers right
- **Spend totals come from `documents`.** Spend per supplier, establishment, month or period is `SUM(documents.total_amount_ht)`: it is what the Eqolux app shows. Use `orders` (one row per product line) for product-level detail: what was bought, quantities, prices, categories, origins. Its sum can be lower than document totals when some lines of a document could not be read; if you show both, explain the difference.
- **Credit notes are already negative** in totals (`total_amount_ht`, `quantity`, `total_weight_kg`, `co2_kg`), so a plain `SUM` nets them out. Never change their sign. For price statistics (min, max, average, change over time), exclude them: `WHERE NOT is_credit_note`.
- **Compare like with like.** For food (`products.item_category_type = 'food'`), compare `price_per_kg`; for other products, `unit_price_ht`. When asked "the price of X" for a food product, give the price per kg first. For an average price, divide spend by weight or quantity (`SUM(total_amount_ht) / SUM(total_weight_kg)`) rather than averaging line prices.
- **Currency.** Amounts are already converted to the display currency (EUR unless `display_currency` says otherwise). Never convert them yourself. State the currency in the answer.
- **Dates** are UTC calendar days. Filter with date literals or `CURRENT_DATE`, for example `WHERE o.date >= CURRENT_DATE - INTERVAL '12 months'`.
- **Default period.** When the user gives none, use the last 12 months and state the exact dates used.
- **Drafts.** `is_draft` marks data the customer has not validated yet. The app includes it; include it too, unless the user asks for validated data only.
- **Data quality.** `error_risk` is higher for lines read with low confidence. When a price looks implausible (an extreme price per kg, a unit mix-up), check `error_risk` and `unit` and say so instead of presenting it as a fact.
## Finding things by name
- **Products:** match `products.name`, `name_fr` and `name_en` with `unaccent(lower(...)) LIKE '%term%'`; try the singular, the plural and the other language. If nothing matches, the user may mean a supplier: search `organizations` where `type = 'Supplier'`.
- **Establishments:** `organizations` of type `Client`, `Client Sub` or `Group`. Groups often stand for regions or countries ("our Swiss hotels").
- **Categories:** `item_categories` is a tree (`parent_id`); a product's categorPreview of the first 4,000 characters. Read full instructions →
Referenced files: 2 — file references are metadata, not package contents.
--- name: eqolux-savings-opportunities description: Find savings in restaurant or hotel purchasing with Eqolux — the same product bought at very different prices, price increases, purchases above a negotiated fixed price. Use when the user asks where they could save money, why costs went up, which prices increased, or wants to prepare a supplier negotiation. license: MIT metadata: version: '0.1.0' author: Eqolux --- # Savings opportunities in Eqolux data Apply the `eqolux-purchasing-analysis` rules first: pick the organization with `list_organizations`, read `describe_data`, exclude credit notes from price statistics, compare food per kg and other products per unit, and link records through `app_url`. Then look for the three kinds of savings below. `references/savings-queries.md` holds a tested query for each; adapt the period and thresholds to the user's request. ## 1. The same product at very different prices For each product over the period: lowest and highest price paid, spend, and the saving if every purchase had been made at the lowest price, that is `spend - lowest price × volume`. Keep products with meaningful spend (by default at least 1,000 in the display currency) and a highest-to-lowest ratio of at least 1.3. Give the documents where the lowest and highest prices were paid, so the user can check them. ## 2. Price increases Compare each product's average price, weighted by volume, between two periods: by default the last 6 months against the 6 months before, or the same months a year earlier when the user cares about seasonality. Rank by extra cost: `(recent price - earlier price) × recent volume`. Ignore products bought in only one of the two periods. ## 3. Purchases above a negotiated price `products.fixed_price` holds a negotiated price, in the unit given by `fixed_price_unit` (`kg` or `unit`). Compare it with the price actually paid and rank by extra cost: `spend - fixed price × volume`. ## Presenting the findings - Show the top 10 by estimated amount, with product, supplier, prices, volume and the estimate, each product and document linked through `app_url`. - Call every figure an **estimate**, and say how it was computed and over which period. - Point out what could distort it: a unit mix-up (kg against pieces), lines with a high `error_risk`, a product that changed packaging, seasonal products. - End with the two or three actions most worth taking, for example the supplier to call first and what to ask for. - On a very large organization, a query over 12 months can time out: narrow it to a category, a supplier or a shorter period.
Referenced files: 2 — file references are metadata, not package contents.
Publisher declarations from the archived package. These are separate from our research and the live service's terms.
Package observed Nov 21, 2026.
plugin_asdk_app_6ac60b5bb9908191b155794b75a13fc7
Download plugin data (JSON)Plugin details: freshly prepared. Prepared . Refreshed when plugin data changes or after 6 hours. Your account and Follow status are live.