Local-first desktop workflow

Keep the signal. Cut the tab sprawl.

JobScraper pulls public company boards into one dense PyQt workbench, stores everything in SQLite, and gives you a fast place to filter, inspect, export, and plan the next skills to level up.

Watch the workflow Run it locally

Public-source first, with direct ATS boards and curated job feeds.
Dense desktop layout with companies, jobs, and selected-job detail side by side.
Background exports, source imports, diagnostics, analytics, and roadmap generation.

310 bundled public sources in the current seed

25k job scale target for performance work

SQLite local workspace with exports, backups, and reports

PyQt desktop-first UI built for high-volume scanning

What the shipped app focuses on

One workbench for browsing, validating, and narrowing real openings.

The public release is intentionally centered on no-auth sources and local operator control. You can inspect source health, edit source rows, export filtered job slices, reopen the getting-started flow, and clean up generated files directly from the GUI.

Local workspace

Settings, sources, logs, reports, backups, and exports are visible to the user and not hidden behind a cloud account.

Honest scope

The bundled public release does not ship login-backed LinkedIn, Indeed, or browser-only portals as default sources.

Unified workbench

Browse companies, jobs, and normalized detail without context switching.

The main surface keeps the company scope, job table, and selected-job description on one screen. Search, stack, source family, source row, tag, HN review mode, and filter presets all stay in the active browsing loop.

Company sidebar with matching and open counts
Job table with stack, location, source, and posted timestamps
Selected-job detail with normalized metadata and outbound links

JobScraper workbench with company filters, a job table, and selected job detail.

Source diagnostics

See which sources are healthy, blocked, new, or failing.

The Sources tab exposes raw source rows, health groups, quality score, last scrape status, source focus, and watchlist probing in one operational view.

Health filters with focused diagnostics
Focus In Workbench for source-scoped browsing
Edit Source and Probe Watchlist actions

Analytics and roadmap

Turn the current job slice into scope-aware signals and study direction.

Analytics and Topic Roadmap stay lazy-loaded, disclose their active scope, and help you understand what the current market slice is asking for.

Scope labels, sample size, and refreshed-at context
Selected-jobs, current-filters, or whole-DB roadmap modes
Local heuristic roadmap generation from stored job text and stacks

JobScraper topic roadmap tab showing scope controls and generated learning guidance.

Workspace controls

Keep paths, exports, and generated files visible to the operator.

The app exposes runtime paths, concurrency settings, Local AI fields, export workflows, the startup tutorial, and a storage cleanup surface from the Tools menu.

Import preview with backup before source mutations
Background export with cancellation and final destination feedback
Storage visibility for logs, reports, backups, exports, and caches

JobScraper getting started dialog explaining the first-run workspace flow.

Real app captures

Walk through the current product surface.

The JobScraper workbench showing company selection, jobs, and selected job detail. — The core browsing surface keeps company counts, job rows, and selected detail aligned on one screen.

The source diagnostics tab showing health filtering and source metadata. — Sources lets you sort out healthy, new, blocked, or failing rows before they waste time during a scrape.

The topic roadmap tab with selected-job scope and generated study checklist. — Roadmap uses the active scope to turn repeated demand signals into a concrete study checklist.

The getting started dialog for new JobScraper users. — The first-run guide introduces import, scrape, filtering, export, source health, and storage cleanup without blocking normal work.

Job-seeker flow

How the app is meant to be used.

Load or edit your public sources, then run a scrape.
Use companies, source family, source row, tag, stack, and text search to narrow the slice.
Inspect job detail, export the filtered set, and use roadmap output to decide what to study next.

JobScraper browsing workflow on a populated workbench.

In progress

Upcoming features worth watching.

These are the live follow-up areas already tracked in the local roadmap and source audit files.

Adapter work

Broader public-board coverage

Fortinet, Hudson River Trading, and the Juniper / HPE networking split still need either a verified no-auth feed or a dedicated adapter path.

Discovery work

Deeper aggregator intake

AIJobs export-feed validation and HiringCafe company-intake flow are still open follow-ups after the first aggregator pass.

Source promotion

More verified source rows

The watchlist and HiringCafe appendix still contain promotable public boards once additional direct probes or new adapters land.

Quick start

Pick the build that matches your machine.

Windows and Linux ship as packaged desktop builds. macOS currently runs from source. Every path still writes its local workspace under Documents/JobScraper.

Packaged build

Windows

Portable EXE plus installer flow for a normal desktop install.

JobScraper.exe inside the packaged app folder
JobScraperSetup.exe when the installer build is present

Download Windows build

Packaged build

Linux

PyInstaller onedir bundle with launcher and icon assets for desktop environments.

JobScraper executable inside dist-linux/JobScraper/
JobScraper.desktop and icon assets for launcher install

Download Linux build

Source build

macOS

There is no packaged .app release yet. Run the PyQt app directly from source today.

Python 3.11+
Editable install with .[build,test]

Build macOS from source

Build from source on GitHub

github.com/d3v4shish/JobScraper

git clone https://github.com/d3v4shish/JobScraper.git JobScraper-Public
cd JobScraper-Public/JobScraper
python -m venv .venv
source .venv/bin/activate
python -m pip install -U pip
python -m pip install -e ".[build,test]"
python -m jobscraper

What ships today

Windows packaged builds from build/build_windows.ps1 with EXE and optional installer output
Linux packaged builds from build/build_linux.sh with executable, desktop launcher, and icon
macOS support through the GitHub source run until a packaged app build lands
settings.json, sources.json, and source_watchlist.json under the workspace config directory
A SQLite database, rotating logs, exports, reports, and backups
Tools actions for import preview, export, activity, logs, reports, watchlist access, tutorial, and settings