Everything inside OptiResume

Recruiters, and the ATS that pre-screens for them, scan for the words in their job description. Paste any JD and OptiResume produces a brand-new variant of your resume that speaks that JD's language: it leads your summary and skills with what the role prioritises, and rewrites bullets to use the JD's terminology wherever the underlying skill is genuinely yours. Each variant is saved as its own row, so you can keep one finely-tuned resume per company or drive instead of endlessly overwriting a single file.

Not every requirement in a JD carries the same weight. This breakdown reads a job description and separates the true must-have skills from the nice-to-haves, reads the seniority signal (intern / fresher / junior / mid / senior / lead), tags the domain, and lists 3-5 of the day-to-day responsibilities in plain English. It then labels each keyword you're missing as either a real blocker or a bonus, so you spend your limited prep time on what actually gates the application.

Anyone can type 'React' on a resume. The Skill Claim card checks whether the skills in your Skills section actually show up in your projects or experience, and flags the ones that don't, because that's exactly the gap a recruiter probes in an interview. You can close a flagged skill two ways: add a bullet that demonstrates you used it, or attach real evidence (a GitHub repo, a certificate, a hosted project, or a live demo). A proof is owned by your account, so one link covers every resume variant.

Your ATS-readiness score (0-100) is computed from 27 deterministic local rules, no language model sits anywhere in the formula, so the same resume always produces the same number. Every rule checks one concrete thing (contact parseability, section headers, quantified bullets, date formats, and so on) and returns pass / warn / fail with a one-line fix. Tap 'How is this scored?' on any result to see the entire breakdown, weights and tier badges included, plus the industry-standard 40/40/20 pillar view (JD-match / parseability / content).

Every one of the 27 scoring rules maps to published evidence: recruiter surveys, ATS parsing benchmarks, a randomized field experiment on how skill listings are read, and government career guidance. The engine scores three things that evidence actually supports, machine readability, recruiter findability, and human-skim quality, and skips the folklore entirely.

Paste a JD from any job posting, including Naukri and Internshala (give us those URLs and we'll fetch the posting for you), and we extract its keywords, then show precisely which ones your resume covers versus misses, with a 0-100 match score. The matching is smart, not literal: 'ML' counts as 'machine learning', 'k8s' as 'Kubernetes', and a semantic-similarity layer catches conceptually-equivalent skills the alias map misses, so 'Generative AI' correctly matches 'Large Language Models', marked with an ≈ badge.

After every analysis we ask a single question: 'Did this resume get you a callback?' Your answer is anchored to the exact score values of that analysis. In aggregate, these answers let us calibrate the engine's rule weights against real hiring outcomes from students and early-career candidates, the difference between scoring on heuristics and scoring on evidence.

Enter your academics once and instantly see whether you meet the hard gates, CGPA, 10th/12th %, active backlogs, gap months, branch, for 25+ major recruiters: TCS NQT, Infosys, Wipro, Accenture, Cognizant, Capgemini, HCLTech, LTIMindtree, Deloitte, and more. Each drive returns a color-coded verdict (green = clears / amber = borderline / rose = auto-reject) with a per-gate breakdown and a confidence label, checked against the latest published criteria.

If you're from a non-CS engineering branch aiming for a software role, the analysis adapts: we auto-detect your branch category and, when it's non-CS/IT, surface a Branch Advice panel with reframing tips tuned to your background, how to position lab work, mini-projects, and coursework so recruiters read 'engineer who can code', not 'off-branch applicant'.

Weak bullets are the quiet killers of a fresher resume. This rewrites them into the STAR pattern (Situation-Task-Action-Result), leading with a strong verb, adding a concrete result, and folding in metrics where your career stage honestly supports them. Each rewrite comes with a one-line rationale so you can see what changed and why.

Goes a level above individual lines: a recruiter's-eye pass over the whole resume covering structure, ordering, gaps, framing, and what's conspicuously missing for your target role. You get 3-5 prioritised recommendations, each with a severity tag, so you know what to fix first.

Paste the messy description of a college, hackathon, or personal project, even a single informal line, and get back a polished project entry: a specific title, a canonical tech-stack list, and 3-5 STAR-compliant bullets with measurable outcomes, ready to drop straight onto your resume. Thin input yields honestly-short output; it never pads with fabricated detail.

Generates a focused cover letter from your resume plus the target company, role, and optional JD context. It opens with a specific hook that pairs this company and this role with your strongest credential, and stays anti-fabrication-guarded throughout, your resume is the only source of facts. It also hands you a ready-to-send recruiter email, a subject line plus a tight intro, that you can copy or open straight in your mail app.

After you've scored against a job description, one tap generates 5 to 8 interview questions tailored to that role: technical questions from the JD's real requirements, and behavioural and project questions from your own resume. Each one comes with what the interviewer is really probing for and an answer hook, pointing at the exact project or skill on your resume to anchor your reply. At least one question is designed to pressure-test a specific claim you made, because that's what a real interviewer does.

A lightweight tracker for your whole job hunt: log applications across companies and statuses (Applied / OA / Interview / Offer / Rejected / Withdrawn), and each entry auto-links to the exact resume version and analysis you used for it. Set next-action dates so a follow-up or an OA deadline never quietly slips.

Turn any resume, including a JD-tailored variant, into a properly formatted document you can actually submit: a single-column PDF or Word (.docx) file with real text, clear section headings, and clean bullets. No tables, columns, text boxes, or images, the exact layout that survives an ATS parser intact. The bullets even render as vector marks so a machine reads the text cleanly while a recruiter sees a crisp dot.

Three connections that turn separate tools into one cockpit. One-tap 'Track this application' files a tracker entry straight from a report with the resume, analysis, and JD already linked, no retyping. Each tracker card then shows that resume's ATS score as a readiness chip, so you can see at a glance you applied to Razorpay with a resume that scored 61, and tailor before you apply. And 'What changed since your last check' diffs your latest analysis against the previous one, keyword gaps closed, checks improved with their point gains, weak bullets fixed, so you can see your edits actually paid off.

Generate a public, branded share card of your ATS score for LinkedIn, Twitter, or WhatsApp. Your display name is auto-redacted to first name + last initial for privacy. Share it, paste back your post URL, and earn +3 AI credits once the post is verified.

Every analysis can mint a share badge: a high-resolution square image with your score ring and verdict, stamped with a QR code. The QR resolves to your badge's short link (/badges/<code>), which lands on your public cert page showing the verified score, JD match, and top matched skills, with your name shortened to first name + last initial. Anyone who scans it sees the live number straight from OptiResume, not a screenshot.

A curated ladder of 100 must-do programming problems, ten themes of ten, ramping from FizzBuzz and string reversal up to two-sum and Kadane's, ending exactly where the coding rounds at TCS, Accenture, Wipro, and Tech Mahindra live. Each problem opens to a plain-language statement, two worked examples, and a hint that names the technique without spoiling the solution. Tick them off as you solve them (anywhere you like, your editor or any judge), watch your progress ring fill, and hit milestones along the way, Quarter Century, Half Century, the Nervous Nineties. Finish all 100 and you earn a serious, verifiable certificate.

Complete all 100 problems and you're conferred the title of Code Centurion on a genuine certificate: deep-navy and gold, your name, the date, a unique code, and a QR that resolves to a public verification page. Anyone who scans it, a recruiter, a classmate, lands on a page confirming it's real, along with your journey (when you started and how long it took), an honest signal that rushed 'completions' can't fake. Download it as an image and post it with the ready-made LinkedIn caption.

A hand-picked, filterable library of the resources that actually get freshers placed, split into Technical Knowledge (DSA sheets and courses, practice platforms, CS fundamentals, system design, languages, web, AI/ML, tools) and Aptitude (MCQ banks, video channels, books, and the current 2025-26 test patterns for TCS NQT, Infosys, Wipro, Accenture, Cognizant, Capgemini, and Tech Mahindra). Every card credits its creator by name and links only to the official source, we curate and point, we never copy their content.

Our own original quick-recap cards for the night before a test or interview: 12 aptitude topics (percentages, time-speed-distance, probability, and more), each with the formula, one worked micro-example, and the classic trap to avoid, plus flip-style Q-and-A decks for OS, DBMS, Computer Networks, and OOP built from the questions freshers actually get asked. Every topic is designed to be skimmed on a phone in about three minutes.

One place to see your remaining AI credits, understand what consumes them (bullet rewrites, suggestions, project polish, cover letters, interview-prep packs, resume tailoring, and never the free rule-based score, exports, tracker, or prep), request more directly from Yash, and grow your balance through the Share & Earn loop.

Edit your display name and phone number from Settings. A unique-phone rule quietly prevents one person from farming multiple accounts for free credits, keeping the free tier fair for genuine students.

Under the hood: 15-minute access tokens, refresh-token rotation on every renewal, and a hard 8-hour session ceiling that never moves. If a token is ever reused (a theft signal), the entire session family is revoked automatically. Logging out kills the whole family server-side, not just your local state.

Delete your account whenever you want from the Danger Zone. It's a soft-delete: we keep aggregated, de-identified data for analytics, but you can no longer log in, and your email and phone remain reserved so they can't be re-registered by someone else.