GetPacked (@getpackedai)

NEW HERE? LET ME SHOW YOU AROUND GETPACKED. 𝗔𝗟𝗚𝗢𝗥𝗜𝗧𝗛𝗠𝗦 — The classics—but “Accepted” isn’t the finish line. Every core problem continues with 1–3 follow-ups that force you to adapt. 𝗔𝗟𝗚𝗢𝗥𝗜𝗧𝗛𝗠𝗦 𝗜𝗡 𝗧𝗛𝗘 𝗪𝗜𝗟𝗗 — Algorithms don’t live only in interview questions. Ever seen a hash map as your game’s…

SQL MAKES MORE SENSE WHEN YOU CAN SEE THE TABLE CHANGE. SQL Lab on mobile is a table-first learning experience built to show how a query transforms data, clause by clause. Use it to: • Inspect the starting rows and columns. • Follow filtering, grouping, joining, sorting, and aggregation as they…

“GOOD. NOW WHAT IF WE CHANGE THE RULES?” That sentence is where memorized solutions usually fall apart. While we provide real coding problems on the Desktop version, we wanted to show a version that could work when you sign in in your phone to continue practice GetPacked Follow-Ups take a base…

ALGORITHM PRACTICE, BUILT FOR STUDYING ON A PHONE. While we provide real coding problems on the Desktop version, we wanted to show a version that could work when you sign in in your phone to continue practice. GetPacked’s mobile Algorithm Problems focus on the part that fits the screen well…

CAN YOU DEBUG IT IN YOUR HEAD? Circuit on mobile gives you compact code-reasoning cases you can work through without needing a full coding setup. Practice across algorithms, debugging, backend systems, and SQL as you: • Trace a short piece of logic. • Predict the result or identify the failure…

SHORT DOES NOT HAVE TO MEAN SHALLOW. Quick Drills are touch-first exercises built for the moments when you have a few minutes and want a real mental rep—not another piece of content to scroll past. You may be asked to: • Predict what the code returns. • Put steps in the correct order. • Spot the…

GETPACKED MOBILE VERSION: KEEP THE THREAD ALIVE. You do not need a marathon study session every time you open GetPacked. Daily Practice gives you a fresh rotating mix of focused questions you can finish from your phone. Expect practice across areas such as: • Coding patterns. • Python behavior…

PRACTICE DOES NOT ALWAYS NEED A LAPTOP. GetPacked Practice Home brings focused study modes into a phone-first space, so you can choose the kind of practice that fits the time and attention you have right now. From Practice Home, you can jump into: • A fresh daily practice session. • Quick drills…

THE ENDPOINT FAILED. THE BUG MAY NOT BE IN THE ENDPOINT. Backend failures travel across layers. A wrong import, a broken dependency, an async mistake, or a bad data assumption can all surface as the same ugly response. Debugging: Backend Edition trains you to follow that trail. You will…

THE DATABASE IS NOT MAGIC BEHIND THE MODEL. GetPacked’s SQLAlchemy practice helps you understand what the ORM is doing—and what you need to tell it to do correctly. Work through exercises involving: • Modern model definitions and column behavior. • Sessions and transaction boundaries…

WATCH IT RUN — SEARCH IN ROTATED SORTED ARRAY One confident mistake can sink this one: Strict less than on the sorted-half check. THE PLAY 1. One half of the array is always sorted after rotation. 2. If nums[left] at most nums[mid], the left half is sorted. Check whether target lies in…

WATCH IT RUN — SEARCH A 2D MATRIX For Search a 2D Matrix, the hard part is refusing this mistake: treating each row independently with O(m log n) instead of leveraging the cross-row sorted property to get O(log(m*n)). THE PLAY 1. Because every row's first element is larger than the previous row's…

WATCH IT RUN — MEDIAN OF TWO SORTED ARRAYS The entire problem turns on one rule: Searching the shorter array bounds the work by min(m,n), the key to beating an O(m+n) merge. THE PLAY 1. Always binary search on the shorter array to guarantee O(log(min(m, n))) time. 2. If you partition array A at…

WATCH IT RUN — KTH SMALLEST ELEMENT IN A SORTED MATRIX Kth Smallest Element in a Sorted Matrix looks simple until returning mid at the end. THE PLAY 1. You are not searching for an index, you are searching for a value. The search space is [matrix[0][0], matrix[-1][-1]]. 2. To count elements at…

WATCH IT RUN — FIND PEAK ELEMENT One confident mistake can sink this one: Comparing nums[mid] with both neighbors. THE PLAY 1. Because nums[-1] and nums[n] are both negative infinity, the array is guaranteed to have at least one peak. 2. If the middle element is less than its right neighbor, the…

WATCH IT RUN — FIND MINIMUM IN ROTATED SORTED ARRAY Find Minimum in Rotated Sorted Array looks simple until comparing nums[mid] against nums[left]. Track the invariant, not the choreography. THE PLAY 1. The minimum sits exactly at the rotation point, where the sorted order breaks. 2. Compare…

WATCH IT RUN — BINARY SEARCH The green check depends on one detail: Wrong loop condition. THE PLAY 1. Halve the search space every iteration. That's where O(log n) comes from. 2. Use left at most right as the loop condition (closed interval). When nums[mid] less than target, set left = mid + 1…

WATCH IT RUN: MEETING ROOMS II 🏢 That min-heap is NOT storing meetings because it loves calendars. It only cares about one thing: When does each occupied room become free? Sort meetings by start time. For each meeting, compare its start with the earliest end time at the top of the heap. start >=…

WATCH IT RUN: BEST TIME TO BUY AND SELL STOCK 📈 This algorithm is not predicting the market. Relax, Warren Buffett. It remembers the cheapest price you've seen BEFORE today. For each price: min_price = min(min_price, price) profit_today = price - min_price best = max(best, profit_today) The…

WATCH IT RUN - ROTATE IMAGE A 90° clockwise rotation becomes much easier when you split it into two geometric moves. First transpose the matrix across its main diagonal. Then reverse every row. The diagonal flip followed by the horizontal flip produces the rotation in place, with O(1) extra space…

WATCH IT RUN - INTERSECTION OF TWO LINKED LISTS Different starting lengths stop mattering when both pointers walk the same total distance. Let one traverse A then B and the other traverse B then A. Both paths are len(A) + len(B), so the offset cancels and the pointers align at the shared node - or…

WATCH IT RUN - COIN CHANGE "Take the biggest coin first" sounds reasonable. It is not reliable across coin systems. Let dp[a] mean the fewest coins needed to build amount a. Set dp[0] = 0, leave unreachable amounts at infinity, and try every coin that can lead into each amount: dp[a] = min(dp[a]…

WATCH IT RUN - CONTAINER WITH MOST WATER The widest container is where the search begins. The shorter wall decides where it goes next. Area is width × min(left_height, right_height). Moving the taller wall only shrinks the width while the shorter wall still limits the height, so it cannot improve…

WATCH IT RUN - GROUP ANAGRAMS Anagrams stop being a string problem once every word gets a fingerprint. Sort the letters in each word, use that sorted form as a hash-map key, and every matching word falls into the same bucket. That is the invariant: the same letters always produce the same…

WATCH IT RUN - LONGEST REPEATING CHARACTER REPLACEMENT The whole window is controlled by one equation: window_size - max_frequency <= k Keep the most common character and replace everything else. If the number of required replacements exceeds k, move the left edge until the window is valid again…

WATCH IT RUN - SPIRAL MATRIX Four boundaries can replace an entire visited grid. Walk the top row, right column, bottom row, and left column; shrink the matching boundary after each sweep. Once the boundaries cross, the work is done. The extra guards before the bottom and left passes prevent…

WATCH IT RUN - MAXIMUM SUBARRAY Kadane's algorithm is one decision repeated: extend the current subarray or start fresh here. For each value, set current = max(value, current + value), then compare that with the best total seen anywhere. A negative history is not worth carrying forward when the…

WATCH IT RUN - REVERSE LINKED LIST The hard part is not flipping an arrow. It is avoiding the loss of everything after it. Save next, point curr.next to prev, then advance prev and curr. Those three pointers preserve the unprocessed list while reversing one link at a time. The invariant: prev…

WATCH IT RUN - TWO SUM "Use a hash map" is not an explanation. The reason it works is that the map remembers every value we have already passed. For each number, compute the complement needed to reach the target. If that complement is already in the map, the pair is done. If not, store the current…