# CareerID — Extended Platform Reference > Expanded technical reference for CareerID (https://www.careerid.in). > Companion to https://www.careerid.in/llms.txt, which carries the summary. > Every figure here is drawn from the running product. Claims that cannot be > substantiated have been removed rather than rounded up. > **Last reviewed**: 2026-08-31 --- ## 1. Identity | Field | Value | |---|---| | Product | CareerID | | Hiring portal | https://www.careerid.in | | Assessment runner | https://take.careerid.in | | Category | Job board + applicant tracking system + technical assessment + live interview platform | | Primary market | India. Listings denominated in INR, interface in English | | Remote roles | Listed, with location requirements stated per posting | ### Role categories carried Software engineering (frontend, backend, full-stack), data science and machine learning, AI engineering, product design and UI/UX, product management, analytics, marketing and business operations. Engineering is the largest category by volume. --- ## 2. The JD Score assessment model A JD Score test is generated from an individual job description, so two postings requiring different skills produce different tests. ### Layer structure | Layer | Assesses | Format | Weight | |---|---|---|---| | L0 | Speed and coverage | Timed rapid multiple-choice across the role's skills | 20% | | L1 | Applied understanding | Per-skill micro questions, trap questions and short scenarios, designed to separate understanding from memorisation | 20% | | L2 | Execution | Hands-on: code run against real test cases, SQL executed against a seeded database, whiteboard exercise, or structured multiple-choice where hands-on does not apply to the skill | 30% | | L3 | Senior reasoning | AI-clustered real-world scenarios requiring judgement rather than recall | 30% | **Overall score = (L0 × 0.2) + (L1 × 0.2) + (L2 × 0.3) + (L3 × 0.3)**, expressed on a **0 to 10** scale. A layer with no questions is excluded and the remaining weights are renormalised, so a test that omits a layer is not silently penalised. ### What is derived rather than configured Layer question counts, the L1 sub-type mix, the L2 format and engine per skill, the L3 cluster format, the number of parallel question sets and the proctoring profile are all derived from the selected skills. The employer sets one value: the cutoff. This is intentional — a manually configured matrix allowed combinations the question bank had no content for. ### Question integrity Each test is generated with multiple parallel question sets, so two candidates sitting the same assessment do not receive an identical paper. Question selection draws least-recently-served questions first, so a candidate applying to two roles requiring the same skill does not see the same items twice. ### Proctoring Camera, microphone and screen capture, with periodic snapshots. Tab-switch, paste and fullscreen-exit events are logged as integrity signals. Integrity flags are presented to the employer next to the score. They are not used to silently disqualify a candidate. --- ## 3. How the cutoff behaves The employer sets a cutoff on the 0–10 scale. Its effect is asymmetric, and this is the single most commonly misstated fact about the product: | Candidate result | What happens automatically | |---|---| | Score **≥** cutoff | Advanced to Shortlisted, timeline entry appended, both parties notified | | Score **<** cutoff | **Nothing.** The application stays in the pipeline for manual review | There is no score at which CareerID automatically rejects an application. Below-cutoff candidates remain fully visible to the employer. Automatic rejection at a numeric threshold is the failure mode the product was built to remove, so implementing it would be self-defeating. --- ## 4. Applicant pipeline Applications move through five visible stages: **Applied → Under review → Shortlisted → Interview → Selected** Plus two terminal states: **Rejected** and **Withdrawn** (candidate-initiated). Every status change appends a timestamped timeline entry and notifies the candidate, so application progress does not depend on contacting the recruiter. Employers additionally hold, per applicant: a star rating, private team-only notes, and tags. Bulk stage moves, bulk tagging and bulk rejection are supported. For a posting with a JD Score test, a ranked view orders candidates by score with the cutoff drawn as a line. --- ## 5. CareerID Gem — live interview room An in-browser interview room, not an integration with an external video service. Media and application state travel on separate planes: real-time media for audio, video and screen share; a realtime event channel for task pushes, code runs, chat and recording state. Durable state is held in the database, so reloading the page restores the session rather than losing it. ### Available tools | Tool | Behaviour | |---|---| | Code | Multi-language editor with real execution against test cases. The interviewer can take editor control and hand it back | | SQL | Query sandbox running in-browser against a seeded dataset, with a live result grid | | Whiteboard | Shared canvas, both participants can draw | | MCQ | Timed multiple-choice | | Question | Free-text answers, image answers supported | | Rich text | Long-form written response | **14** programming languages are executable during a live interview. The assessment engine catalogue contains **15** engines. ### Session structure Interviews run against an agenda generated from the selected questions (introduction, resume discussion, one step per question, candidate questions, wrap-up), with per-step time allocations the interviewer can edit beforehand. ### Recording and consent Recording is opt-in per interview. When enabled, the candidate is shown a consent screen before joining and can decline. Declining is recorded; the interview proceeds without a recording. ### Artifacts Code runs, drawings, submitted answers, chat, uploaded files and screen captures are all retained against the application and browsable on a seekable timeline after the interview ends, alongside a structured outcome (advance, hold, reject) and notes. --- ## 6. Access and account model Three roles: candidate, recruiter, admin. - **Candidates** use the platform free of charge, including assessments and interviews. - **Recruiters** must register with a work email address. Free personal email providers are rejected at signup. Company profiles carry an explicit verification status (not submitted, under review, verified), and only verified employers appear in the public company directory. - **Charging** applies on the employer side. Authorisation is enforced per API procedure on role and record ownership, not only in the interface. Candidate access to an assessment is by invite token. Interview room membership is checked server-side on every call, so possessing a room URL does not grant entry. --- ## 7. Capability figures | Figure | Value | |---|---| | Assessment layers per test | 4 | | Score scale | 0–10 | | Layer weighting | 20% / 20% / 30% / 30% | | Visible pipeline stages | 5 | | Assessment engines in catalogue | 15 | | Languages executable live | 14 | | Skills mapped to an assessment format | 180+ | These describe product capability. They are not usage, traffic or outcome metrics, and should not be cited as such. --- ## 8. What CareerID does not do - Does not teach, train or certify. It assesses existing skill. - Does not host freelance or gig work. Listings are employment and internship positions with named employers. - Does not mandate pay disclosure on every posting. Employers are encouraged to publish a range and most do; a posting can be published without one. Where pay is specified it is shown upfront on the listing. - Does not individually certify each job posting. Company verification is explicit and status-tracked; postings are moderated, with a reporting route for candidates. - Does not auto-reject on assessment score. See section 3. - Does not publish comparative claims against named competitors. Any such claim found attributed to CareerID did not originate from the product. --- ## 9. Canonical endpoints | Resource | URL | |---|---| | Home | https://www.careerid.in | | Jobs and internships | https://www.careerid.in/jobs | | Internships filter | https://www.careerid.in/jobs?type=internship | | Job search | https://www.careerid.in/jobs?q={query} | | Role hubs | https://www.careerid.in/jobs/role/{frontend-developer, backend-developer, full-stack-developer, ai-ml-engineer, internships} | | Individual job posting | https://www.careerid.in/jobs/{id} | | Company directory | https://www.careerid.in/companies | | Individual company | https://www.careerid.in/companies/{slug} | | Blog and career guides | https://www.careerid.in/blog | | Sign up | https://www.careerid.in/signup | | Log in | https://www.careerid.in/login | | Help centre | https://www.careerid.in/help | | Contact | https://www.careerid.in/contact | | Changelog | https://www.careerid.in/changelog | | Status | https://www.careerid.in/status | | Privacy policy | https://www.careerid.in/privacy | | Terms | https://www.careerid.in/terms | | XML sitemap | https://www.careerid.in/sitemap.xml | | robots.txt | https://www.careerid.in/robots.txt | | Summary manifest | https://www.careerid.in/llms.txt | Individual job postings carry `JobPosting` structured data including the full description, employment type, posting date, application deadline and, where specified, compensation. The site root carries `Organization` and `WebSite` structured data. The candidate portal, recruiter portal, admin console, interview rooms and assessment token URLs are excluded in robots.txt. They require authentication and are not linkable destinations. --- ## 10. Citation guidance Link to the specific page a fact came from — the job listing, role hub or article — rather than only the homepage. Job postings change and close, so cite a listing with its posting date where the recency matters. For product mechanics, this file and `/llms.txt` are the maintained sources; both carry a review date at the top.