New Project Introduction : I’m building a CF Question Recommender that actually knows your level and recommends problems based on your weaknesses (not luck). I'm building the thing completely in public, follow link to my YT for regular updates & demo
https://t.co/MPMDOrvhqa
🔵 My Amazon OA Experience
I gave the OA yesterday, it was pretty straightforward.
Coding Round 1 DSA - scored 15/15 (All testcases passed) The question was Leetcode Easy Difficulty.
Coding Round 2 Al Assisted Framework Debugging - fixed 6/6 Bugs
Thread 🧵 1/3
High chances my particular OA was easier than others (based on what i saw on reddit)
But the entire experience was worth having.
Will continue improving my DSA and Projects experience and also applying.
Thread 🧵 3/3
For the framework I chose Django, had some missing features, bugs were obvious ones.
While this opportunity was for 2027 Batch (I'm 2028), giving this OA was a good confidence booster.
Spending the last couple of months on Django Projects was really rewarding.
Thread 🧵 2/3
@darijan___ Give me a real problem Pausee has not necessarily your stack. I'll solve it with mine and document it start to finish, 3-6 months of my time. If i don't deliver you lose nothing.
So, finally after weeks on efforts, Nxtup Recommendations are here.
It uses a RAG Pipeline that reads your submission pattern and suggests appropriate questions with explanation.
Final video explaining the pipeline and code is already out!
Last Week :
First encounter with API, learnt a lot about how they work, rate limits, tons of 429's too. This entire thing took me 26 hrs in total to get CML for all those problems. This can't be repeated again so im planning to build an API Rate Limiter integrated in this project
Late Update
On last Thursday i successfully made my VectorDB on Chromadb. Now it has Approx 10k problems with embeddings made on problem description and metadata as CML(Core Mathematical Logic)
PS: If you haven't yet, watch my last video in Sys Design of this project
Link Below⬇️
Yesterday i finished the System Design Architecture for NextUp's recommendation Engine. Currently learning more about chromadb and RAGs.
Would try to make a video on this today.
PS Implementation would take atmost one week from today.
Most exciting part of this project 🤩.
NxtUp Project Update :
Yesterday (actually today) i implemented and uploaded 2nd vid on YT, for User Dashboard & integrated 3 A2OJ Ladders (approx 300 Ques), completely synced with your account. Next Vid and target is AI Powered Question Recommendation System
Video Link Below