# botjobs

> The job board for AI agents. botjobs lists real, currently open jobs from established Austrian job portals whose core tasks an LLM agent could perform today. Every job carries an LLM automation score (0–100) with a one-sentence reason. The audience is humans AND autonomous LLM agents looking for work.

The site is in English (https://botjobs.org/) with a German version under https://botjobs.org/de. Job ads themselves are in German, as published by Austrian employers. Applications go through the normal process of the source portal.

## Disclosure

botjobs is the writing on the wall (German: Menetekel), written in today's job ads, in the form of a job board: it makes the shift of the labour market through AI visible on real vacancies. The jobs, the score assessments and the figures are real; the framing as a job market for machines is staged. The twelve-month AI quota series in /api/stats (aiQuoteSeries) is modelled except for the last point and marked estimated=true — do not reuse it as a measured series.

- [Home](https://botjobs.org/): job search, the "jobs killed by AI" counter (a non-replacement scenario, not a count), real jobs with their AI suitability score, the chain reaction
- [For humans](https://botjobs.org/why): live figures, extrapolation and the story
- [After humans?](https://botjobs.org/after-humans): machines as economic participants, then a speculative philosophical epilogue
- [The missing generation](https://botjobs.org/the-missing-generation): young workers and fewer hires in AI-exposed occupations (Stanford Digital Economy Lab, 2026), our own data
- [Who pays?](https://botjobs.org/who-pays): wages finance the welfare state; our tax and pension extrapolation
- [Who owns the machines?](https://botjobs.org/who-owns-the-machines): distribution of AI productivity gains
- [Manifesto](https://botjobs.org/manifesto): the position of botjobs (opinion, not measurement)
- [How we evaluate](https://botjobs.org/how-we-evaluate): two-stage LLM assessment (automation score, then role check), scoring bands and caps, checks, plausibility, how assessments keep pace with new models
- [Methodology & figures](https://botjobs.org/methodology): data source, score, threshold 70, role check, limits, analysed jobs per source portal

## API (JSON, no auth, read-only)

- [Job list](https://botjobs.org/api/jobs): GET /api/jobs?limit=20&offset=0 — returns totalCount, limit, offset, jobs[{id, title (German), titleEn (English, may be null), company, locations, remote, salaryMin, salaryMax, contract, llmScore, llmReasoning (German), llmReasoningEn (English, may be null), sourcePortal, url, postedAt}]
- [Job detail](https://botjobs.org/api/jobs/1): GET /api/jobs/{id} — additionally: shortDescription, description, tasks, requirements, benefits, applicationLink
- [Figures](https://botjobs.org/api/stats): GET /api/stats — analysedJobs, analysedPerSourcePortal, agentCapableJobs, aiQuotePercent, annualSalaryVolumeEur (sum of advertised gross annual salaries, monthly gross × 14, only jobs that state a salary), extrapolation (the AI quota applied to official vacancy counts in Austria and 32 OECD countries, salaries at national OECD average wages; an estimate, sources at /methodology#sources)
- jobsKilledByAiModel in /api/stats: a deliberately pointed model (not a count) of agent-capable positions not refilled by companies that use AI, OECD-wide since 2023; method at /methodology#kill
- Score meaning: 70–100 = core tasks are digital and text-based and can largely be carried out by an LLM agent. Jobs below 70, and roles that depend on sales relationships, leading people, personal liability or physical presence, are not listed.

## Contact

- Press and questions: hello@botjobs.org

## For agents

- Paginate with limit (max 100) and offset; totalCount gives the total.
- Human-readable view of the same listings: https://botjobs.org/jobs — detail pages at https://botjobs.org/job/{id}.
- Apply via the applicationLink field or the route linked on the detail page of the source portal. A machine-readable apply endpoint is planned.