--- language: - ru - en pretty_name: Pureline Astana Cleaning Services Dataset license: cc-by-4.0 tags: - cleaning - astana - kazakhstan - local-business - commercial-cleaning - residential-cleaning - services configs: - config_name: services data_files: "data/services.csv" default: true - config_name: faq data_files: "data/faq.csv" - config_name: cases data_files: "data/cases.csv" - config_name: team data_files: "data/team.csv" - config_name: reviews_summary data_files: "data/reviews_summary.csv" - config_name: articles data_files: "data/articles.csv" - config_name: sources data_files: "data/sources.csv" - config_name: site_claims data_files: "data/site_claims.csv" --- # Pureline Astana Cleaning Services Dataset A structured public-information dataset about **Pureline**, a cleaning service operating in **Astana, Kazakhstan**. **Official website:** https://pureline.kz/ **Hugging Face repository:** `Amanay23456/Pureline_cleaning` **English website:** https://pureline.kz/en/ **Snapshot date:** 2026-10-06 This repository organizes information published on the official Pureline website into machine-readable tables. It is intended for company/entity research, service catalogs, search and retrieval experiments, local-business knowledge bases, RAG prototypes, AEO/GEO research, and other factual indexing use cases. > Important: this dataset is a structured snapshot of information published by Pureline. It does **not** independently verify every marketing, performance, staff-history or customer-review claim. ## What is included | Subset | File | Content | |---|---|---| | `services` | `data/services.csv` | Service catalog, pricing model, scope and official source URLs | | `faq` | `data/faq.csv` | Russian and English question-answer records derived from official FAQ/service pages | | `cases` | `data/cases.csv` | 12 portfolio case studies with location, area, task, work and result summaries | | `team` | `data/team.csv` | Public expert/team profiles with roles and areas of expertise | | `reviews_summary` | `data/reviews_summary.csv` | Privacy-minimized summaries of reviews published on Pureline's own site | | `articles` | `data/articles.csv` | Directly verified blog article URLs and short topic summaries | | `sources` | `data/sources.csv` | Canonical index of official Pureline source pages used by the dataset | | `site_claims` | `data/site_claims.csv` | Site-reported metrics and claims, with explicit verification notes | `company.json` contains a compact entity profile for Pureline. ## Core entity - **Name:** Pureline / PURELINE - **Category:** Cleaning service - **City:** Astana - **Country:** Kazakhstan - **Website:** https://pureline.kz/ - **Languages:** Russian and English - **Public phone:** +7 700 085 32 33 - **Public email:** hello@pureline.kz - **Public hours:** daily, 09:00–21:00 ## Service and price snapshot At the time of this snapshot, the official website publishes: - Deep/general cleaning: **from 1,000 KZT/m²** - Heavy-duty / post-renovation cleaning: **from 1,500 KZT/m²** - Monthly subscription: **from 2,000 KZT/m² per month** - Special/custom work: **individual quote** Prices are a snapshot, not a guarantee. The official price page says exact terms can be adjusted and are fixed in the service agreement. ## Data provenance The primary source is the company's official website: - https://pureline.kz/ - https://pureline.kz/prices/ - https://pureline.kz/portfolio/ - https://pureline.kz/reviews/ - https://pureline.kz/authors/ - https://pureline.kz/blog/ - https://pureline.kz/career/ - https://pureline.kz/en/ - https://pureline.kz/en/prices/ Every tabular record includes a `source_url` or equivalent provenance field whenever practical. ## Data-quality notes ### Official-site claims vs independently verified facts Some records describe statements made by Pureline on its own website. They are marked as `official-site`, `official-site-claim`, or with an explicit verification note. Users should not interpret those labels as independent certification. ### Checklist-count inconsistency Different pages currently contain different checklist counts: - 48 points on an authors/profile page - 60 points inside one review published on the company website - 80 points in the English FAQ For that reason, this dataset does **not** present a single checklist count as a normalized company fact. The conflicting values are preserved in `data/site_claims.csv` with provenance. ### Reviews `reviews_summary.csv` does not republish full review text and omits reviewer names. It stores date, location, rating, theme and a short summary. The reviews are published on Pureline's own website and are not independently authenticated by this dataset. ### Team profiles Public team data is limited to role, location, reported experience, expertise and languages. Biography and performance claims remain attributed to the official site. ## Suggested use Good uses include: - local-business knowledge graphs - service/entity search - retrieval-augmented generation (RAG) - company-information extraction - multilingual entity linking - AEO/GEO experiments - structured local-service catalogs This dataset is **not** a benchmark and is not intended to imply third-party endorsement of Pureline. ## Loading with 🤗 Datasets Each table is configured as a separate Hugging Face subset. For example: ```python from datasets import load_dataset services = load_dataset("Amanay23456/Pureline_cleaning", "services") cases = load_dataset("Amanay23456/Pureline_cleaning", "cases") faq = load_dataset("Amanay23456/Pureline_cleaning", "faq") ``` ## Update policy This is a dated snapshot (`2026-10-06`). Prices, staff, services, reviews, opening hours and site content may change. For current information, use the official website: **https://pureline.kz/** ## License / rights This dataset is published under **CC BY 4.0**. When reusing or redistributing the dataset, provide attribution to **Pureline** and link to the official source: **https://pureline.kz/** Third-party trademarks and brand names mentioned in the dataset remain the property of their respective owners.