Welcome to the Jungle Jobs Scraper & Hiring Signals

Search Welcome to the Jungle (WTTJ) jobs, export full descriptions and review the wording behind technology mentions and hiring themes. Filter recent or remote jobs, focus on company pages and reuse a previous run to deliver only job IDs you have not observed before. French and English rules are included.

By Monkey D. Scraping. Get full descriptions, source links and reviewable evidence. Up to 100 inspected jobs per run.

Price: $0.003 per run start + $0.0005 per delivered job, including platform usage and full details. Ten jobs cost $0.008; 100 cost $0.053. The start charge also applies to empty searches and may apply to failed runs. There is no extra full-details fee. The minimum selectable maximum-run budget is $0.004; this is a spending ceiling, not a minimum charge. With a lower run budget, fewer jobs are inspected; RUN_REPORT.budget shows the budget applied.

中文用户 / 中文用戶： 简体中文说明 · 繁體中文說明（Welcome to the Jungle职位采集、欧洲/法国招聘数据、招聘信号与销售线索）

First result

{"query":"data engineer","maxJobs":10,"countryCodes":["FR"]}

This inspects up to ten unique jobs, newest first, and returns their full descriptions. Open the dataset for jobs, RESULTS.csv for a spreadsheet-safe export and RUN_REPORT for run details. Salary and application-link fields are null when the listing does not publish them.

To limit the search to an employer, supply companyUrls, for example https://www.welcometothejungle.com/fr/companies/doctolib, or "companySlugs":["doctolib"]. To retrieve an employer's jobs without a keyword constraint, set query to an empty string. Use WTTJ company pages or slugs, as plain URLs without embedded filters. Query, company, country, date and remote filters combine. Companies and countries within each list are OR alternatives.

For recent remote roles, add "publishedWithinDays":7,"remote":"fulltime". Zero days means any publication date. The date window is fixed at run start in UTC and checked again against the full details. Use ISO country codes such as GB.

What the output means

Fields	Meaning
id, jobUrl, applyUrl	Source identity, original listing, published application URL when available
description, candidateProfile	Full normalized text, kept separate
company, companySlug, companyWebsite	Employer context from the source
locations, remote, contractType, language	Published location and employment context; language is source metadata
salaryMin, salaryMax, salaryCurrency, salaryPeriod	Original salary units, as published
publishedAt, updatedAt, observedAt	Publication, source update and our observation, kept distinct
technologyEvidence, technologyMentions	Terms observed in the description, with exact supporting text and matchedText
signals, signalTypes	Explicit wording about team expansion, a first/founding role or a system migration
sourceTools	Separate source-provided taxonomy
contentHash	Stable hash for comparing the same record in your own workflow; excludes retrieval time
observationType	export, initial, or first_seen; refers to this Actor's observation, while publishedAt gives the publication time

Every evidence item contains field, start, end and excerpt. Offsets are UTF-16 positions in the normalized description: JavaScript description.slice(start,end) reproduces the excerpt. Long sentences are shortened around the match and marked excerptIsPartialSentence. Team expansion covers employees and contractor pools. A first/founding role refers to a team, market or project.

Hiring themes use conservative French and English phrase rules, not an LLM or a confidence score. Negations, candidate-experience wording and selected hypotheticals are filtered out, so each theme rests on explicit wording. Jobs in other languages are exported in full. An empty signals array means no explicit theme wording was found.

technologies annotates jobs rather than filtering them. Omit it to use a curated list including Salesforce, HubSpot, Snowflake, Databricks, Kubernetes, Terraform, AWS, Azure, Google Cloud, BigQuery, dbt, PostgreSQL, Kafka, SAP, ServiceNow, Microsoft Fabric, Power BI, Fivetran and Redshift. The common spelling PowerBI maps to Power BI, with the actual matched text preserved. An explicit empty list disables technology matching.

Only jobs not observed before

1. Run with "mode":"new" and no baseline. The initial sample is returned as observationType: initial.
2. Open RUN_REPORT.nextRunInput. It contains the same settings and the key-value store ID to use as baselineStoreId next time.
3. Run that input to return only IDs missing from the selected baseline. The report shows inspected and skipped counts.

A successful all-results run also provides a reusable baseline. Use a baseline from a completed run of this Actor with the same query, countries, companies, date window and remote filter; you may change the inspection limit or technology annotations. The storage picker requests only READ access to the selected store.

Baseline memory holds up to 10,000 job IDs, and store retention follows your Apify settings. Known IDs count toward the inspected jobs without another detail request, and are skipped even if their content later changes.

Example of reviewable evidence

One private canary found a system-migration theme in Funding Circle's Data Platform Engineer listing:

{"type":"system_migration","excerpt":"We are moving to Snowflake and to a federated \"Data as a Product\" model."}

The actual record also includes the source URL, full description and exact offsets.

Three ready-to-use examples

Search jobs in France

{"mode":"all","query":"data engineer","maxJobs":10,"countryCodes":["FR"]}

Find founding-role wording in GB-listed jobs

{"mode":"all","query":"founding","maxJobs":10,"countryCodes":["GB"]}

Initialize a new-jobs observation

{"mode":"new","query":"data engineer","maxJobs":10,"countryCodes":["FR"]}

For the next observation, copy RUN_REPORT.nextRunInput, including its baselineStoreId.

Technical review covered 100 real descriptions, balanced between French and English; those cases are kept as regression fixtures. Every theme carries its supporting text, so you can review it before using it in an automation.

Ready-to-run examples

Example	Use case	What you get
WTTJ hiring signals: French startups hiring Account Executives	Sales and recruiting leads	20 Account Executive jobs published in France in the last 30 days, mostly from SaaS startups, with company, location, contract, remote policy and full description

Browse all ready-to-run examples →

Practical uses

Recruiters can review role descriptions and original application links. Technology consultants can inspect the exact sentence behind a technology mention. Researchers can compare published salaries without mixing annual and monthly units. Export CSV or JSON, or consume the standard Apify dataset API in an existing workflow.

Cost and implementation

Pricing is per run start and delivered job as described above. A short browser bootstrap is used only when the public page cannot be read directly; job search and details use public HTTP endpoints afterward. Public search configuration is discovered on each run and never embedded in source, output or input. No WTTJ account, personal cookies, external credentials or paid proxy is required by the implementation.

Chinese Simplified

中文说明：Welcome to the Jungle职位采集与招聘信号工具（Welcome to the Jungle Jobs 爬虫 / API）

Welcome to the Jungle Jobs Scraper & Hiring Signals 按关键词、公司、国家、发布时间和远程办公类型，采集 Welcome to the Jungle（WTTJ） 上的公开职位：完整职位描述、候选人要求、薪资、申请链接和公司信息。同时标注职位描述中提到的技术（如 Salesforce、Snowflake、AWS），并识别团队扩张、首个/创始岗位、系统迁移等招聘信号，每条都附带原文摘录，方便核对。支持法语和英语规则，适合获取法国及欧洲招聘数据。

为什么选择这款WTTJ职位采集工具？

• 完整职位描述：返回规范化后的完整描述和候选人要求，两者分开保存，并附原始职位链接和公开的申请链接。
• 可核对的招聘信号：团队扩张、首个/创始岗位、系统迁移等信号基于保守的法语和英语短语规则，不依赖大模型；每条信号都附带原文摘录和精确位置。
• 技术提及附原文：默认技术列表涵盖 Salesforce、HubSpot、Snowflake、Databricks、Kubernetes、Terraform、AWS、Azure、Google Cloud、BigQuery、dbt、PostgreSQL、Kafka、SAP、ServiceNow、Microsoft Fabric、Power BI、Fivetran 和 Redshift，也可以自定义最多30个技术词。
• 只看新职位：使用“仅新职位”模式，以上一次运行为基准，只返回之前没有见过的职位ID。
• 薪资保持原始单位：保留原始的最低/最高薪资、币种和周期，不做猜测性的年薪换算，便于准确比较。
• 无需账号：无需WTTJ账号、个人Cookie或付费代理。
• 导出方便：数据集可导出为JSON、CSV、Excel，另附 RESULTS.csv 表格文件和 RUN_REPORT 运行报告，也可以通过Apify API接入你的系统。

适用场景

• 招聘与人才研究：按关键词和国家查看法国、英国、德国等地的最新职位，阅读完整描述并直接打开申请链接。
• B2B销售线索与技术咨询：找出在招聘中提到 Snowflake、Salesforce、Kubernetes 等技术的公司，并查看提及该技术的原句，人工核实后再跟进。
• 招聘信号监控：关注正在扩张团队、招聘首个岗位或进行系统迁移的公司。
• 重点公司跟踪：输入公司页面或公司标识（如 doctolib、qonto），获取这些雇主在WTTJ上的职位。
• 薪资研究：比较公开的薪资区间，月薪和年薪不会混在一起。
• 定期获取新职位：配合“仅新职位”模式，每次只拿到新出现的职位。

可以获取哪些数据？

• 职位信息：职位ID、职位链接、申请链接、完整描述、候选人要求、合同类型、工作地点、远程办公类型、语言。
• 公司信息：公司名称、公司标识、公司官网。
• 薪资：最低薪资、最高薪资、币种、薪资周期（保持原始单位）。
• 时间：发布时间、来源更新时间、本工具的采集时间，三者分开记录。
• 技术提及：technologyMentions 和 technologyEvidence，含原文摘录和匹配文本；来源网站自带的工具分类单独保存在 sourceTools 中。
• 招聘信号：signals 和 signalTypes，包括团队扩张、首个/创始岗位和系统迁移，每条附原文摘录和位置。
• 对比字段：contentHash 用于在你自己的流程中比较同一条记录，observationType 标明记录来自普通导出、初始样本还是首次见到。

价格

按次启动 + 按交付职位计费，平台用量和完整详情均已包含：

内容	价格
每次运行启动	0.003美元
每个交付的职位	0.0005美元
示例：10个职位	0.008美元
示例：100个职位	0.053美元

完整详情不另收费。启动费在搜索结果为空时同样收取。可在Apify控制台设置每次运行的最高费用（最低可选0.004美元），这是费用上限而不是最低收费。

使用方法

1. 在 query 中输入搜索关键词，例如 data engineer。
2. 在 countryCodes 中选择国家代码（如 FR、GB、DE），留空表示所有国家。
3. 设置 maxJobs，即本次要检查的职位数量（最多100个），按发布时间从新到旧。
4. 可选：用 companyUrls 或 companySlugs 指定公司；用 publishedWithinDays 限定最近几天发布的职位；用 remote 筛选远程办公类型（如 fulltime）。
5. 点击 Start，完成后导出为JSON、CSV、Excel，或通过API获取。

{
  "mode": "all",
  "query": "data engineer",
  "maxJobs": 10,
  "countryCodes": ["FR"],
  "publishedWithinDays": 7,
  "remote": "fulltime"
}

只想获取某家公司的全部职位时，将 query 设为空字符串并填写 companySlugs，例如 ["doctolib"]。

常见问题

需要Welcome to the Jungle账号或Cookie吗？ 不需要。工具读取WTTJ的公开职位页面，无需账号、个人Cookie或付费代理。

如何只获取新发布的职位？ 第一次运行时设置 "mode": "new"，返回初始样本并建立基准。之后打开 RUN_REPORT.nextRunInput，其中包含相同的设置和下一次要使用的 baselineStoreId，用这份输入再次运行，即可只返回之前没有见过的职位ID。

招聘信号是怎么识别的？ 基于透明的法语和英语短语规则，而不是大模型或置信度评分。每条信号都附带原文摘录和在描述中的精确位置，你可以随时核对。

可以自定义要识别的技术吗？ 可以。在 technologies 中填写最多30个技术词；不填则使用默认列表。技术识别只做标注，不会过滤职位。

可以通过API、自动化工具或AI智能体调用吗？ 可以。支持Apify API、Python和Node.js客户端，以及n8n、Make、Zapier等自动化工具；也可以通过Apify MCP服务器作为AI智能体的工具使用。

中文说明会翻译采集到的数据吗？ 不会。中文说明只是使用文档；采集到的数据保持来源的原始语言（职位描述按发布时的原文呈现）。

其他工具

• 招聘信号情报工具：采集 Greenhouse、Lever 和 Ashby 上的公司招聘页面，提供附原文的技术提及、扩张信号、职级、地点和公开薪资范围。
• Google趋势关键词分析工具：查看关键词和技术名称的Google搜索热度历史。
• 网站SEO审计与联系方式采集工具：审计目标公司的公开网页，获取SEO问题、联系方式和安全响应头信号，适合代理商开发客户。
• Google广告采集工具（Google广告透明度中心）：按域名或品牌查看公司在Google搜索、YouTube和Google购物上的广告。

Chinese Traditional

中文說明（繁體）：Welcome to the Jungle職缺擷取與招募訊號工具（Welcome to the Jungle Jobs 爬蟲 / API）

Welcome to the Jungle Jobs Scraper & Hiring Signals 依關鍵字、公司、國家、發布時間與遠端工作類型，擷取 Welcome to the Jungle（WTTJ） 上的公開職缺：完整職缺說明、應徵條件、薪資、應徵連結與公司資訊，並標示說明中提到的技術，以及團隊擴編、首位／創始職位、系統遷移等招募訊號，每筆都附原文摘錄。支援法語與英語規則。

為什麼選擇這款WTTJ職缺爬蟲？

• 完整職缺說明：說明與應徵條件分開保存，附原始職缺連結與公開的應徵連結。
• 可核對的招募訊號：以保守的法語與英語片語規則判斷，不依賴大型語言模型，附原文摘錄與精確位置。
• 只看新職缺：「僅新職缺」模式以上一次執行為基準，只回傳之前沒看過的職缺ID。
• 不需帳號：不需WTTJ帳號、個人Cookie或付費代理。

適用場景

• 招募與人才研究：依關鍵字和國家查看法國、英國、德國等地的最新職缺。
• B2B銷售名單：找出職缺中提到 Snowflake、Salesforce、Kubernetes 等技術的公司，人工核實後再跟進。
• 重點公司追蹤：輸入公司頁面或公司代稱（如 doctolib、qonto）。

使用方法

1. 在 query 輸入搜尋關鍵字，在 countryCodes 選擇國家代碼（留空代表所有國家）。
2. 設定 maxJobs（最多100個）；選填 companySlugs、publishedWithinDays 與 remote。
3. 點擊 Start，完成後匯出JSON、CSV、Excel，或透過API取得。

{
  "mode": "all",
  "query": "data engineer",
  "maxJobs": 10,
  "countryCodes": ["FR"],
  "publishedWithinDays": 7,
  "remote": "fulltime"
}

價格

每次執行啟動0.003美元，每個交付的職缺0.0005美元，平台用量與完整詳細資料皆已包含：10個職缺0.008美元，100個0.053美元。搜尋結果為空時同樣收取啟動費。

常見問題

需要Welcome to the Jungle帳號或Cookie嗎？ 不需要。工具讀取WTTJ的公開職缺頁面。

如何只取得新發布的職缺？ 第一次執行設定 "mode": "new" 建立基準，之後使用 RUN_REPORT.nextRunInput 再次執行即可。

招募訊號怎麼判斷？ 依透明的法語與英語片語規則，而非大型語言模型；每筆訊號都附原文摘錄。

中文說明會翻譯擷取到的資料嗎？ 不會。說明是中文，資料保留來源的原始語言（職缺說明以刊登時的語言呈現）。

完整欄位與輸出範例請見英文說明或简体中文说明。

Related Actor

Hiring Signal Intelligence covers company boards on Greenhouse, Lever and Ashby. This Actor focuses specifically on Welcome to the Jungle jobs and their supporting text.