Engineering insights, algorithmic problem-solving techniques, open-source maintainer guides, and systems architecture authored by students and mentors at BIT Mesra.
An engineering deep-dive into how distributed databases guarantee linearizability across network partitions: Paxos tradeoffs, Raft mechanics, log compaction, and scaling with Multi-Raft groups.
A deep architectural investigation into how modern LLM inference runtimes like vLLM, TensorRT-LLM, and SGLang achieve high throughput—from memory-bound decoding and KV cache fragmentation to virtual memory paging and continuous iteration-level scheduling.
A deep dive into how algorithmic intuition is developed—moving beyond memorizing algorithms to identifying invariants, problem reductions, and constraint signals.
Mastering tree query techniques for ICPC and Codeforces Div 1: a rigorous guide to Heavy-Light Decomposition for path queries and Centroid Decomposition for divide-and-conquer path counting.
How we eliminated latency and memory bottlenecks in student-scale microservices by migrating from centralized Node.js/Go containers to zero-copy streaming Web Streams on global V8 edge isolates.
A practical blueprint for students wanting to contribute to production codebases without being overwhelmed by massive repositories.
How ACM BIT Mesra approaches architectural simplicity, static generation, and edge hosting to build student platforms that never crash during registrations.
