Feature · Cross-language search

Cross-language Telegram search for selected sources

Ask in English across selected non-English Telegram sources, retrieve conceptually related posts, and inspect the original-language evidence.

Direct answer. The Ocean lets a researcher ask an English question over indexed sources written in another language and receive an English synthesis with citations to the original messages. Critical names, modality, dates, and idioms still require source-level review.

Research fit

Who this workflow is for

English-speaking researchers who monitor selected public sources in Russian or other working languages.

01

Describe the concept

Ask with entities and events rather than depending on a single machine-translated phrase.

02

Retrieve across wording

Use semantic retrieval while retaining exact spellings, aliases, and date boundaries where known.

03

Review the original

Open every material citation and ask a fluent reviewer to inspect high-stakes wording.

Practical module

Cross-language review record

Preserve the path from English question to original evidence.

English questionThe neutral research question and the output language requested.
Language variantsOriginal spellings, aliases, transliterations, and domain-specific vocabulary.
Reviewer noteAny uncertainty involving negation, modality, sarcasm, names, or relative dates.

Cross-language search prompt

The template keeps original-language verification in the workflow instead of hiding translation uncertainty.

Copy-ready query

Answer in [output language] using [selected source languages] posts from [period] about [entity or event]. Include alternative spellings, cite original posts, preserve uncertainty and negation, and flag passages that require a fluent reviewer.

Inputs to set

Languages
Question language, source languages, and desired answer language.
Variants
Known original spellings, abbreviations, aliases, and transliterations.
Review risk
Terms, claims, or entities where translation precision is especially important.

Expected output

  • English synthesis linked to original-language posts
  • Key original wording and bounded translation
  • Uncertainty and fluent-review queue

Review before use

  • Preserve negation and modal verbs
  • Resolve names and relative dates manually
  • Do not infer that translation coverage is exhaustive
Download Markdown resource
Read the complete resource
# Cross-language search prompt

The template keeps original-language verification in the workflow instead of hiding translation uncertainty.

## Query

Answer in [output language] using [selected source languages] posts from [period] about [entity or event]. Include alternative spellings, cite original posts, preserve uncertainty and negation, and flag passages that require a fluent reviewer.

## Inputs

- **Languages:** Question language, source languages, and desired answer language.
- **Variants:** Known original spellings, abbreviations, aliases, and transliterations.
- **Review risk:** Terms, claims, or entities where translation precision is especially important.

## Expected output

- English synthesis linked to original-language posts
- Key original wording and bounded translation
- Uncertainty and fluent-review queue

## Review checks

- [ ] Preserve negation and modal verbs
- [ ] Resolve names and relative dates manually
- [ ] Do not infer that translation coverage is exhaustive

Boundaries

What this does not establish

  • Machine-generated synthesis can lose nuance or strengthen tentative wording
  • Language support does not replace a fluent reviewer for critical findings
  • Only selected indexed sources are searched

Founding research pilot

Apply with a cross-language search task

Bring one recurring research job and a small set of public sources. We will assess fit personally before offering a pilot.

Apply for a 14-day research pilot

30 founding researcher places · rolling admission · personal reply within 2 business days

Continue researching

Related guides

Publication notes. This page was reviewed before publication and is scheduled for review by 2026-12-19. Facts used: product-cross-language · product-citations · product-hybrid-search · product-selected-sources