We use cookies for site analytics. Accept to help us understand how the site is used. See our Privacy Policy for details.
Features and improvements shipped on gitGood.dev - 82 of them, newest first. We ship weekly.
Every job on the Openings board now has a 'Prep for this interview' link - one click carries the role into the prep-plan builder, prefilled from the listing. Paste the full posting for an even sharper plan.
Try it →In the week before your interview, the dashboard countdown opens a kit built for the day itself: your own STAR stories condensed to a page you can skim, the questions to ask them for each round of that company's loop, what they say they value, and a 20-minute warm-up. Print it, or add the whole run-up to your calendar.
Try it →Paste a job description and get the full study plan before you sign up - two per month, no account needed. Create a free account afterwards and the plan you built comes with you and stays on your dashboard.
Try it →Signed in, the Job Market Pulse now opens with the news that matters to you - ranked by the companies you have applied to, your upcoming interview, the prep packs you have opened, and your target track - with the reason on every item.
Try it →Pick a theme on any story in your bank and get the 90-second answer you would actually say, with delivery notes and the follow-up to expect. Behavioral resume questions now open a pre-tagged story too. Your learning-path certificates show on your public profile, all your stars have a page at /starred, and previous attempts can be diffed against the reference solution.
Try it →Switch languages, close the tab, come back tomorrow - your code is still in the editor. Drafts are kept per problem and per language on this device, and the AI code review now stays on screen after you close the Accepted dialog.
Try it →The Java and Go language pills now actually run and submit your code against the real JDK and Go toolchain sandboxes. If you tried them since the 18th and got "Validation failed", that was us - fixed.
Try it →The timed exam simulator now keeps your answers and the clock if the page reloads mid-exam, warns before you leave, and its time limit is set from each real exam's pace. The free 5-question sampler now spans every domain of the exam instead of the first one.
Try it →Microphone access for voice mode was being refused by a server header - not by your browser - and is now allowed. Streaming replies in mock interviews, salary coaching and AI-enabled practice no longer drop words, get stuck, or hide the reason when you hit a daily limit.
Try it →Write the 5-8 stories you'll actually tell, tag them to the behavioral themes they answer, and see which themes you have no story for yet. Ask the coach for a critique - per-section scores, rewrites you can paste in, and the follow-ups an interviewer will ask. The behavioral mock interview now reads your bank and probes your real stories.
Try it →Upload your resume and get the 10-15 probing questions a careful interviewer would ask about it - the gaps, the numbers, the tech you named - each with what they're testing, how to answer, and a practice link. The mock interviewer weaves them into follow-ups.
Try it →Every learning path now tracks what you've done: solves count automatically, reading items get a tick box. Finish a path and claim a verifiable certificate with a one-click LinkedIn 'Add to profile' link.
Try it →Six new frontend problems run inside a live document - createElement, ARIA state, reordering rows, walking text nodes - and are graded on what ends up in the page. JavaScript or TypeScript; find them on the Frontend path.
Try it →Under each problem you'll find the pattern it belongs to and the cue that should make you reach for it, a hint ladder you can open one step at a time, the optimal approach with its time and space complexity, what that problem's own test suite actually checks, and the mistakes that pattern punishes. The hints and the approach stay tucked behind a click, so they're there when you want them and out of the way when you don't.
Try it →Booked an exam? Put the date on your dashboard and pick your cert - you'll get a study plan that re-paces itself as the date gets closer: learn the domains, then drill questions, then run the timed simulator, then cram. Works for all 146 certs we cover, and tells you honestly when one doesn't have a practice exam behind it yet.
Try it →Java and Go join JavaScript, TypeScript and Python in the code editor - real compilers, not simulations, so you get the compiler errors and stack traces you'd get locally, pointing at your own lines. Starter signatures are ready on 295 of our 296 problems. If you interview in Java or Go, you can now practice in it.
Try it →Have an interview on the calendar? Set the date (and company) on your dashboard and get a countdown with a day-by-day focus that shifts as it gets closer - foundations first, then system design, then behavioral and your company's loop, then final polish. Name a company we have a prep pack for and it's woven in.
Try it →Meta now runs coding rounds where you're given an AI assistant and graded on how you use it - and Google is piloting the same. Rehearse the format for real: solve a problem with an AI assistant and a live code executor, then get a scorecard on the collaboration itself - problem framing, AI direction, verification, and code ownership. Free accounts get 20 assistant messages and 2 scorecards a day.
Try it →26 of our 146 certification guides now have a graded practice exam behind them, and the index finally shows which. Every cert card carries a Practice exam badge, and you can filter the list down to just those - on the main index and on each provider's page.
Try it →Every one of the 26 practice banks is now sequenced into at least one role path, so a certification is part of the curriculum rather than something you had to go find. The Data Engineer path gained a certification section it never had; SRE/DevOps, Security Engineer and the three cloud deep-dives all grew.
Try it →The professional and specialty tiers now have practice banks, not just landing pages: AWS SysOps, Solutions Architect Pro, Security Specialty and DevOps Pro; Azure AZ-204, AZ-400, AZ-500 and DP-203; and five GCP Professional exams including Cloud Architect, Security Engineer and Data Engineer. 26 banks and 1,128 questions in total.
Try it →AWS CLI, Azure CLI, gcloud, Helm, GitHub CLI, and a side-by-side comparison of the same operation across all three clouds - including where the equivalence quietly breaks down.
Try it →All 13 banks grew - 405 questions to 608. The eleven that were sitting at 25 are now 40 to 50, so a practice run is long enough to be a real rehearsal and a second attempt is not the same questions again.
Try it →Aimed at new-grad prep: trees and BFS, boundary binary search, bit manipulation, and greedy. Several exist to teach the trap - minimum depth is not the mirror of maximum depth, and the obvious balanced-tree check is quadratic.
Try it →Topological sort, tries, prefix sums and intervals were the thinnest pattern pages - each now has 7-8 problems instead of 2-3. Includes meeting rooms, insert interval, course schedule II, wildcard word search and search autocomplete, all with JavaScript, Python and TypeScript solutions.
Try it →Check an answer and each incorrect option explains itself right where it sits - with your own pick called out - instead of one paragraph you have to search. The timed exam review does the same.
Try it →Every problem is tagged with the pattern that solves it - sliding window, two pointers, BFS, dynamic programming and 26 more. Filter and search on the Code page, or browse a pattern's own page to learn the trigger condition before you drill it.
Try it →Every path now routes you to the SQL problems, quick-reference sheets, and certification practice exams that fit the role - not just MCQs and coding. The data paths gained real SQL practice, and the cloud paths link their matching cert exam.
Try it →AWS Cloud Practitioner and Solutions Architect Associate now run the real exam's 65 questions on the real clock. Plus a new JavaScript category: 40 questions on the event loop, closures, this-binding, and the coercion rules interviewers love.
Try it →No more top-100 cutoff - page through the full board, 30 members at a time, on every period. Every member with an account is on it somewhere.
Try it →Press Ctrl+K / ⌘K anywhere to jump to any problem, topic, or page. Your dashboard and public profile now show a practice heatmap. The code editors gained font size, theme, vim keybindings, and tab-size settings - plus confetti when you earn it.
Try it →Pick from 16 preset avatars or upload a photo in Settings. Your avatar shows on the leaderboard, navbar, and public profile.
Try it →Star any coding problem, SQL problem, system design topic, behavioral theme or cheat sheet and it turns up in ⌘K and on your dashboard. Coding problems keep your previous submissions so you can diff them against what you are writing now, the problem/editor split drags to any width, and zen mode hides everything but the work.
Try it →A first-time correct solve on any practice surface - multiple choice, coding, or SQL - keeps your daily streak alive. Keyboard shortcuts landed on all three surfaces too.
Try it →An email preference centre in Settings, so you can silence one kind of email - streak reminders, the weekly digest, job alerts - without unsubscribing from all of them.
Try it →After a coding solution is accepted, get a staff-level review of your code: complexity, strengths, ranked improvements with before/after rewrites, and an interview-readiness score.
Try it →Paste a JD and get a study plan built from real gitGood content - learning paths, question categories, system design topics, and behavioral themes matched to the role.
Try it →111 curated pair pages compare compensation between companies, level by level, linked from every comp page.
Try it →Two new opt-in emails: matching roles from the openings board as they appear, and a Sunday summary of what you solved, where your streak stands and what to pick up next.
Try it →Three more banks: GCP Cloud Digital Leader (free for members), GCP Associate Cloud Engineer, and Azure AI-900.
Try it →The site now defaults to dark, and the screenshots on the landing page follow whichever theme you are using instead of always showing the light ones.
A second run mode alongside practice: a countdown clock, no feedback until you submit, and a pass or fail at the exam's real cut score - then a full review of every answer.
Try it →Streaks, quests, achievements and your progress charts are no longer behind Premium. Free accounts get the whole dashboard.
Try it →The free trial used to be monthly-only, which quietly pushed anyone who wanted to try the product onto the more expensive plan. Both plans now start with 10 days free.
Try it →Recommendation systems, feature stores, ad serving and real-time bidding, and ML training pipelines - the system design questions that actually come up in ML and AI infrastructure loops.
Try it →A new surface: original scenario-style questions written from the published exam guides - never brain dumps - with per-domain scoring so you can see which domain is weak, and a full explanation on every choice. Launched with AWS CLF-C02, SAA-C03 and DVA-C02 plus Azure AZ-900.
Try it →Behavioral packs for Nvidia, Uber, Airbnb, LinkedIn, Databricks and Anthropic; four more system design topics including LLM inference and RAG; five more cheat sheets (Kafka, GraphQL, Terraform, LLMs and prompt engineering, SQL query tuning); and the PM/TPM case library doubled to 23.
Try it →50 questions on RAG, embeddings and vector search, prompting vs fine-tuning, agents and tool use, evals, inference cost, safety, and transformer fundamentals.
Try it →Pick a company and the AI interviewer role-plays that company's loop - its values, its core topics, its round names - on top of whichever mode you choose: general, behavioral or system design.
Try it →The daily multiple-choice question no longer needs Premium - solve it every day to keep your streak. The coding half completes the pair for the daily leaderboard.
Try it →Solve coding challenges in TypeScript alongside JavaScript and Python, with typed starter signatures and the same sandboxed execution against the real test cases.
Try it →Buy a gift subscription, get a redemption code, send it on. Redeeming grants a full year - useful for a mentee, a graduating student, or a team member.
Try it →Verify a .edu address and get 20% off Premium for as long as you keep the subscription.
Try it →Per-seat pricing for cohorts - $5 a seat per month, $4 at 100 seats and up - with a quote calculator and concierge setup rather than a self-serve checkout.
Try it →The referral page now shows who has brought the most people in, so the program has some visible social proof rather than being a silent credit.
Try it →Every tag now has its own page, so 200+ posts are navigable by subject instead of only by recency.
Try it →A public tracker built on the same live data that powers Pulse - who cut, how many, and when.
Try it →Around 181 pages breaking compensation down per company per role ladder, level by level, rather than one page per company.
Try it →High-impact functional bugs earn the reward too, not just security findings - along with a Bug Hunter badge and a place in the Hall of Fame.
Try it →60 to 78, closing the gaps that existed at the top of the ladder: higher solve tiers, difficulty-specific tiers, and achievements for the non-coding surfaces.
Try it →Service-by-service AWS, Azure and Google Cloud paths for solutions architects, backed by 92 new cloud questions - taking the path library to 19.
Try it →The thinnest categories got filled out: Kubernetes, Terraform, DevOps and SRE, Networking, plus A/B Testing, TypeScript, Mobile, Data Engineering, Frontend Frameworks and Object-Oriented Design.
Try it →The Amazon behavioral set is now complete - a STAR answer for every Leadership Principle, not just the popular half.
Try it →A new surface: write real SQL against seeded databases and get graded instantly on the result set. Joins, aggregation, subqueries, CTEs, set operations, window functions and date handling across four schemas.
Try it →A single 0-100 score on your dashboard and public profile that weighs breadth, depth, recency and consistency - and weights them differently depending on the role you are targeting.
Try it →A judgment-based content type rather than multiple choice: product design, metrics, strategy and prioritisation, estimation and execution cases, each with a framework, a worked answer, the common pitfalls and likely follow-ups.
Try it →Their own learning paths, question categories and readiness weighting - so the platform stops assuming every user is preparing for a software engineering loop.
Try it →Two more roles with dedicated curricula, question coverage and a role-aware readiness score.
Try it →Pick a path at signup and earn a badge exclusive to it. The Polymath track is for people who would rather not specialise, and quests now credit progress on every surface rather than only multiple choice and coding.
Try it →16 to 38 companies, including the quant and trading firms - Jane Street, Two Sigma, Citadel, Hudson River Trading - cross-linked from the prep packs.
Try it →Three more role curricula, and the onboarding quiz no longer routes people into a dead end when their role had no path.
Try it →The writing backlog cleared - interview guides, role breakdowns and company loop walkthroughs, all QA'd against the live product.
Try it →Everything you have unlocked is now visible on your shareable profile, not just on your own dashboard.
Try it →Compare a current offer against a target one over four years - vesting schedules, refresh grants, federal and state tax, FICA, and an ISO warning where it applies.
Try it →One multiple-choice question and one coding problem per day, the same pair for everyone, with a leaderboard ranked by finish time.
Try it →Everything you have got wrong comes back on a 1/3/7/14/30-day schedule, one card at a time, so a missed question does not stay missed.
Try it →Claim a username and your profile lives at a readable URL you can put on a CV, instead of an account id.
Try it →A free quick-reference section, opening with Big-O complexity, common interview patterns, system design tradeoffs and Unix essentials.
Try it →Role-targeted curricula that sequence the content that already exists into an order worth following - starting with Backend, SRE/DevOps and ML Engineer.
Try it →Public-source compensation ranges by company, role ladder and level, with the date each figure was last checked and a citation for where it came from.
Try it →Four questions after you register, and you land on a learning path suited to your role and level rather than an empty dashboard.
Try it →Prep packs for Tesla, TikTok, Figma, Notion, Salesforce, Adobe, Discord, Canva, Ramp, Rippling, Linear and Perplexity, plus system design walkthroughs for payments, search autocomplete, notifications, real-time collaboration, analytics pipelines, message queues, consensus, CDN and edge, and load balancer internals.
Try it →