Save
The hiring system adopted AI quickly. Job seekers got left trying to explain themselves to machines they cannot see.
And somehow, the burden keeps landing on the applicant: learn the tools, rewrite the resume, decode the algorithm, and pay for another platform just to participate in a process that should already be fair.
A 2025 paper, “AI Self-preferencing in Algorithmic Hiring,” found that large language models can favor resumes that look like their own output. The researchers reported self-preference bias between 68% and 88%. In simulated hiring pipelines across 24 occupations, candidates using the same LLM as the evaluator were 23% to 60% more likely to be shortlisted than equally qualified candidates using human-written resumes.
That should make people uncomfortable. Not because AI is bad. That is too easy. The real problem is that people start changing how they describe their work just to sound acceptable to a system they do not understand. That is why I built Job Search Terminal: not another AI job search app, but a practical system people can own without adding another subscription between them and a paycheck.
The AI paradox
There is an obvious question here: if AI-generated resumes can create bias, why build a tool that uses AI to help tailor resumes?
Fair question.
The problem is not simply that AI touched the text. The problem is black-box evaluation and people losing control over how their experience gets represented.
Job Search Terminal uses AI to parse, compare, draft, and suggest. But the human stays in charge. You bring your own API key. You review the output. You edit the language. You decide what is true, useful, and worth sending.
This is not AI trying to beat AI. It is a tool for helping people stay legible in a hiring process that is becoming harder to read.
A tool for control, not autopilot
Job Search Terminal is a free, local-first job search dashboard. It runs on your own machine. No account. No cloud database. No subscription.
It helps with four practical parts of the search:
- Local data storage. Resumes, job evaluations, generated drafts, notes, and interview prep stay on your machine.
- Fit evaluation. The tool compares your experience against role requirements and shows matches, gaps, and evidence.
- Targeted drafting. It helps generate tailored resume drafts, outreach messages, and application answers without auto-applying for you.
- Pipeline tracking. It gives you a simple dashboard for tracking applications, statuses, companies, and next steps.
It does not promise interviews. It does not apply to you. It does not know your career better than you do.
The point is smaller and more useful: reduce the repetitive work so you can spend more attention on judgment. What is worth applying to? What is the truthful story? What needs clarification? What should you ignore?
That is the work. The tool just gives you a better place to do it.
Why local-first matters
Job search data is personal. Your resume, work history, goals, notes, and writing style are not just files. They are part of your professional identity while it is still being shaped.
Job Search Terminal keeps that information on your computer. It uses AI, which helps with analysis, comparison, summarization, and drafting, without asking you to create another account or store your search inside another platform.
That local setup is not a technical inconvenience. It is an architectural choice.
There is a small cost to owning your data: you install and run the tool from GitHub. If that is not something you normally do, the setup guide includes a prompt you can paste into Claude Code, Codex, or a similar coding assistant. Ask it to install and run the tool with you.
A little more setup. A little more control. I think that is a fair trade.
How it works
Once the app is running, the workflow is simple:
- Add your AI API key and resume. Upload or create one or more resumes. The system uses them to build a profile of your skills, experience, preferences, constraints, and writing voice.
- Find or add jobs. Scan supported job sources, including companies that use ATS platforms such as Greenhouse, Lever, and Ashby. You can also add jobs manually by pasting the company name, role title, job URL, and description.
- Evaluate fit. Ask the tool to evaluate a role. It shows how the job matches your experience, where the gaps are, what evidence already exists in your resume, and what you may need to clarify.
- Act on the evaluation. Generate a tailored resume draft, create outreach messages, prepare application answers, research the company, and move the role through a simple application pipeline.
- Review everything. Every draft is editable. Every recommendation is reviewable. Every output is something to question.
That last part matters most. I do not trust AI blindly, and I do not think job seekers should either.
Why I made it free
The job market is already hard enough.
People looking for work are tired. They are anxious. They are making serious decisions with incomplete information. Charging them for basic organization, resume tailoring, and application tracking feels wrong when we can build useful tools differently.
Job Search Terminal is free for non-commercial use. It is open for people to try, test, break, question, and contribute to.
I am not presenting it as finished or perfect. It was built quickly. It will get better only if real people use it in real searches and tell me what works, what fails, and what needs to change.
Try it. Share it with someone who is searching. Open an issue if something breaks. Suggest what would make it more useful.
The hiring market is increasingly automated and opaque. We cannot fix the whole system with one tool. But we can build tools that hand some agency back to the applicant.
Key takeaways
- Job Search Terminal is a free, local-first job search dashboard that keeps your data on your machine.
- It uses AI to support human judgment, not replace it.
- The goal is to help job seekers organize the search, tailor materials, track applications, and stay in control.
Project page: https://pavel.ux.business/job-search-terminal/
GitHub: https://github.com/uxdesignlab/job-search-terminal
Featured image courtesy: Pavel Bukengolts.
- The piece argues that most AI job search utilities deal with the wrong problem: they only lower barriers for candidates and perpetuate existing power imbalances.
- It contends that the choice of local-first, people-centered tools is a political position on professional data ownership, not simply a technical decision.
