SShortSingh.
Back to feed

Four ATS platforms expose public JSON APIs that make job scraping unnecessary

0
·1 views

A developer discovered that major applicant tracking systems — Greenhouse, Lever, Ashby, and SmartRecruiters — serve job listings through unauthenticated public GET endpoints, eliminating the need for headless browsers or proxies. Companies like Stripe, Spotify, and Ramp post openings through these platforms, each accessible via a company-specific slug found in the careers page URL. Greenhouse returns full job descriptions in a single response, though its content field requires HTML-entity decoding, while Ashby offers the most structured data including remote status, compensation bands, and plain-text descriptions. Lever and Ashby return flat arrays with minimal extra requests, whereas SmartRecruiters requires paginated calls plus separate per-job detail requests to retrieve full descriptions. The author recommends writing a thin normalizer per provider so downstream code remains unaware of which ATS a given company uses.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Developer Builds Shopify App Store Scraper to Monitor Competitor Reviews and Churn

A developer has published a Shopify App Store scraper on Apify that extracts app details and user reviews, including reviewer store name, country, and how long they used the app. Because Shopify App Store pages are server-rendered, no headless browser is required to collect the data. The tool supports filtering reviews by star rating and date, allowing users to isolate new one- and two-star reviews for any competitor app on a recurring basis. Named merchant reviewers in low-star reviews can serve as warm leads, since those stores are most likely to be looking for alternative solutions. The scraper is priced at $2 per 1,000 reviews and $5 per 1,000 apps, and can be scheduled weekly to maintain a continuous competitive intelligence feed.

0
ProgrammingDEV Community ·

Agent Experience (AX): Why Software Must Now Be Designed for AI Consumers

Software is increasingly being consumed by AI agents — such as coding assistants, autonomous workflows, and browser bots — rather than solely by human users or developers. This shift introduces a third audience that evaluates software not on visual design or branding, but on whether it is understandable, predictable, and trustworthy. Industry observers are beginning to call this design consideration 'Agent Experience' (AX), distinct from User Experience (UX) and Developer Experience (DX). Tools like GitHub Copilot, Claude Code, and autonomous agents built on frameworks such as LangGraph and MCP already interact directly with APIs and codebases without human involvement. As AI becomes a routine consumer of software, designing systems with AX in mind may emerge as a key competitive differentiator for technology products.

0
ProgrammingDEV Community ·

Rust Developer Builds Real-Time Logic Simulator Handling 100,000+ Gates on Laptop

A software developer has built a high-performance digital logic simulator in Rust capable of processing over 100,000 primitive gates in real-time on standard laptop hardware. The project was inspired by Sebastian Lague's logic simulator videos and aimed to overcome the severe performance limitations common in traditional visual circuit simulators. To eliminate the slowdowns caused by deeply nested chip hierarchies, the developer implemented a flattening compiler that converts recursive gate structures into a single, flat array before simulation begins. Parallel execution is applied dynamically based on workload size, while a lock-free write-back mechanism prevents false sharing without the need for costly mutex locks. A cache defragmentation pass further boosts performance by sorting and repacking gates in topological order, ensuring CPU prefetchers can keep frequently accessed data in L1 cache.

0
ProgrammingDEV Community ·

Systems Developer Shares Debugging Lesson: Trust Error Traces Over Your Own Instincts

A developer writing on DEV Community recounted a 20-minute debugging session while building a deployment reverse proxy engine, where a client CLI kept hanging on a fetch call. The root cause was a simple Express.js mistake — chaining .json() onto .send(200), which attempted to send data after the response cycle had already closed. The author argues that computers execute exactly what is written, not what the programmer intends, making error stack traces a reliable and precise diagnostic tool. Three core debugging rules are proposed: trust the stack trace over visual code review, check data shapes at interface boundaries, and never dismiss or hide error output. The piece concludes that every debugged error builds deeper systems knowledge, and that error messages, read carefully, never mislead.

Four ATS platforms expose public JSON APIs that make job scraping unnecessary · ShortSingh