Abhishek — Backend Engineer · Distributed Systems

about

I'm a backend engineer from Bengaluru who enjoys building distributed systems, scalable architectures, and infrastructure that works reliably under concurrency, not just when a demo runs once.

At LTI Mindtree, I work on ASP.NET Core backend development, where I improve test coverage and fix static analysis issues. The role has taught me how to understand, debug, and strengthen code written by others. Outside of work, I like building things from scratch, which is why I've built two production-style systems, Compylr and ShardLink, with a strong focus on reliability, scalability, and clean design.

I graduated with a B.Tech in Computer Science (8.88 CGPA) from Graphic Era University in June 2025. I also enjoy competitive programming and have solved 1,450+ problems on LeetCode, which has helped me build strong problem-solving skills and a solid foundation in algorithms.

Actively open to opportunities

Looking for SDE-1 roles at product companies — backend, distributed systems, infrastructure. Based in Bengaluru, open to relocation.

stack

Java

C#

Spring Boot

ASP.NET Core

PostgreSQL

Redis

Docker

Angular

Git

GitHub Actions

experience

Software Engineer

LTM (formerly LTI-Mindtree)

2025 — Present · Bengaluru

Backend development in ASP.NET Core, building and maintaining REST APIs for an existing production codebase.
Improved automated test coverage across previously under-tested modules through unit and integration testing.
Resolved SonarQube-flagged issues, code smells, maintainability warnings, and duplicated-code violations — improving overall code quality gates.
Working with large-scale manufacturing datasets, writing and optimizing complex SQL queries to support data validation and reporting workflows.

projects

Compylr

Sandboxed Code Execution Platform

source → live demo →

A competitive-programming judge that executes untrusted code safely at scale, Docker sandboxing with hard resource limits, an async submission pipeline that never blocks under load, and a submission lifecycle built to fail predictably rather than silently.

Built

Docker sandboxing per submission, 512MB RAM, 1.5 vCPU, 128 PID cap
Tiered execution engine: local warm-container fast path for C++ with automatic failover to a self-hosted Judge0 instance on Oracle Cloud (OCI)
Async submission queue via Channel<Guid>, bounded to 1,000 concurrent jobs
9-state submission lifecycle with JWT auth (HMAC-SHA256) and OTP email verification
11-entity normalized schema with composite indexing for O(log n) leaderboard queries

Hard parts

Designing a fallback path so a Judge0/OCI outage doesn't take down code execution entirely
Keeping the queue non-blocking under concurrent submission bursts
Enforcing sandbox limits without starving legitimate long-running submissions
Designing 3-tier rate limiting that's strict enough to matter, loose enough to not annoy real users

ASP.NET Core 8Angular 20 PostgreSQLDocker Oracle Cloud (OCI)JWT

ShardLink

Distributed URL Shortener

source → live demo →

A URL shortener built around the problem that actually matters at scale: what happens to your keyspace when a node joins or leaves. Consistent hashing keeps redistribution small instead of catastrophic. ShardLink cuts it from 75% down to 29%

Built

Consistent hashing ring with Base62 encoding for key distribution
Reduced key redistribution from 75% → 29% during rebalancing events
Redis caching layer with hit/miss analytics feeding a React dashboard
GitHub Actions CI/CD, deployed on Render / Neon / Upstash / Vercel

Hard parts

Getting the hash ring's virtual-node distribution actually even, not just "mostly"
Keeping cache and source-of-truth consistent across rebalancing without stale reads

consistent hashing ring — how ShardLink distributes keys across nodes

JavaSpring Boot PostgreSQLRedis React

Abhishek — backend engineer building systems that hold up under load.

Let's talk about a role.