Qaz.Vision
Vision of Kazakhstan through documents and public positions
A unified corpus of documents, theses, and public positions.
Find statements by person, office or topic, with links to sources.
Find a statement and check its source
Choose a topic and period, open a document and compare positions. Each record lists available languages and its source; missing text or translations are marked separately.
Corpus density by year
The historical corpus layer is curated. Percentages describe the shape of the current public snapshot, not the completeness of an external archive.
Source: public coverage.json Open coverage.json
Sections and corpora
Purpose
Qaz.Vision is a platform for reading and analyzing a confirmed public corpus, helping users quickly see priorities, topics, and the language of decisions.
Who it serves
For quickly finding quotes, theses, and supporting wording while preparing stories.
For fast access to official positions, assignments, and related documents.
For monitoring topic areas and preparing analytical products from the corpus.
How the public layer works
- Published documents, speeches, interviews, articles, and confirmed working materials from role and personal corpora.
- A single document card with date, type, languages, source links, and topic navigation.
- Thesis annotation, A–G topic blocks, and links between documents, theses, and positions.
- It is not a complete archive of every media mention, retelling, or external interpretation.
- A missing item does not mean the event did not happen: the document may not yet have been included or confirmed.
- Summary figures help with orientation but do not replace reading the source document and publication context.
- Page counters and JSON reflect the current published snapshot at request time.
- The freshness date records the latest confirmed corpus event, not an instantaneous update of every topic.
- For external links, use the snapshot date and API and data , rather than only the absolute document count.
- API and data for JSON, bulk exports, and machine access points.
- API documentation and OpenAPI for technical integrations.
- Ask for quick search across the published corpus.
Role and personal snapshots
Methodology
Only published materials tied to a source and a specific public role enter the corpus.
Each document gets a card, language versions, document type, and source links.
Documents are split into theses assigned to A–G topic blocks, forming the basis for position navigation.
Coverage and boundaries
Qaz.Vision is a research collection, not an exhaustive legal archive. The status page reports updates and source reconciliation. Historical coverage is selective; annual document counts do not measure a politician's activity. A catalog record does not mean its full text or translation is cleared for publication.
| Year | Documents |
|---|---|
| 2014 | 2 |
| 2015 | 3 |
| 2016 | 2 |
| 2017 | 4 |
| 2018 | 1 |
| 2019 | 27 |
| 2020 | 36 |
| 2021 | 33 |
| 2022 | 256 |
| 2023 | 425 |
| 2024 | 848 |
| 2025 | 891 |
| 2026 | 935 |
| Stored texts by language | Documents |
|---|---|
| Russian | 3454 |
| Kazakh | 2135 |
| English | 284 |
A stored text is not necessarily cleared for display. Until rights are confirmed, only metadata and source links are public. One document may have multiple language versions.
Full machine-readable matrix: coverage.json.
How to read evidence
First-person wording or an exact line from a speech, interview, or address.
An official source directly reports that a politician noted, stated, instructed, or emphasized something.
A description of an event, appointment, or procedure. It supports document checking but is not published as a politician's position.
Corrections and reuse
Include the card URL, disputed wording, and primary-source link. The public issue log is available through GitHub Issues.
Create an issue ↗Rights to source texts belong to their publishers. When using metadata and analytical annotation, cite Qaz.Vision, the document card, and the official primary source. No separate open license is granted for source texts.
Theses are generated automatically and pass structural checks. For material conclusions, consult the full text and official source.