Vishal Jadeja

Software Engineer·Full-Stack Dev·Builder

Building scalable back-end systems and clean front-end interfaces. Shipping real-time products at Glitchover — from WebSocket infra to payment integrations.

Resume Get in touch

About

I build scalable backend systems and real-time applications.

I'm a Full-Stack Software Engineer focused on backend architecture, distributed systems, and real-time infrastructure. Currently working at Glitchover, where I design and build production systems that power competitive gaming platforms.

I specialize in the MERN stack and building high-performance APIs, event-driven systems, and scalable backend services. I enjoy solving complex engineering problems — from optimizing database queries and eliminating race conditions to designing resilient backend systems.

Outside of work, I deep-dive into system design, distributed systems, and scalable architectures, and enjoy solving algorithmic problems on LeetCode.

Work Experience

Architected event-driven real-time tournament system with 25+ Socket.io events delivering sub-100ms bracket, standings, and lobby updates to 1,500+ concurrent WebSocket subscribers across room-scoped channels — eliminating polling and cutting API load by 60%.
Engineered extensible multi-gateway payment microservice using Factory + Strategy patterns across Razorpay, Cashfree, and Stripe — enabling new gateway onboarding with zero changes to existing production code and zero controller-level branching.
Secured webhook ingestion pipeline with HMAC-SHA256 timing-safe comparison (crypto.timingSafeEqual) and idempotent order-status checks, eliminating replay attacks and duplicate wallet credits from payment gateway retry storms across 3 providers.
Reduced MongoDB query latency by 45% by redesigning compound index strategy across transaction and payout collections and implementing financial snapshot denormalization capturing 12 immutable fields at transaction time for audit-compliant multi-currency payout accuracy.
Engineered multi-step financial waterfall with floating-point integrity checksum (±0.02 tolerance) covering forex markup, GST tax splitting (CGST/SGST/IGST), promo codes, and platform fees — catching silent arithmetic drift across high-volume cross-currency payment flows.
Implemented Strategy pattern dispatching 5 bracket-generation algorithms (single-elimination, double-elimination, round-robin, Swiss, Battle Royale) at runtime, with O(N) preliminary-round insertion for non-power-of-2 participant counts and Fisher-Yates O(N) shuffle for unbiased seeding.
Engineered pessimistic cache-update strategy with recursive Immer tree-walker simultaneously patching 3 RTK Query cache layers after match results, eliminating redundant GET requests and maintaining bracket consistency client-side without a full refetch.
Built zero-downtime database-driven feature flag system enabling ops-controlled payment gateway toggling and dynamic fee configuration without redeployment, combined with 4-tier RBAC and IP-level ban detection across all payment admin routes.
Optimized payment confirmation latency by deferring non-critical conversion tracking (Meta Pixel, analytics) via setImmediate and parallelizing multi-collection MongoDB writes with Promise.all, decoupling user-visible response time from third-party API performance.
Led PWA migration from CRA (React 16) to Vite (React 19), reducing build times by 80% and improving cold-start performance, unblocking faster release cycles for the engineering team.

Projects

Mintmark

Live GitHub

A passive-first personal branding platform that turns what you learn into platform-ready content for LinkedIn, X, and Medium — simultaneously. Features a unified activity heatmap across all learning sources, AI-powered content generation via BYOK, and a personal knowledge assistant scoped strictly to your own notes and posts.

Next.js 16TypeScriptReact 19SupabaseTanStack QueryZustandTailwind CSS v4Trigger.devUpstash Redis

Syncify

Live GitHub

A seamless tool that converts YouTube playlists to Spotify using intelligent fuzzy matching. Handles OAuth 2.0 flows for both platforms, gracefully manages API rate limits, and reduced manual playlist migration effort by 95%.

React.jsNode.jsOAuth 2.0YouTube APISpotify APIFuzzy Matching

Amazon Clone

GitHub

User-side e-commerce platform replicating Amazon's core features — JWT authentication, product catalog, shopping cart, and order processing. Built with a mobile-first responsive UI using React.js and Redux, backed by scalable MongoDB schemas with efficient indexing.

React.jsReduxNode.jsExpress.jsMongoDBJWT

Movie Recommendation System

GitHub

A content-based movie recommender that finds similar films using cosine similarity on TF-IDF vectorized metadata — genres, keywords, cast, and crew. Built as an interactive Streamlit app where you pick a movie and instantly get five tailored recommendations.

PythonStreamlitScikit-learnPandasTMDB APICosine SimilarityTF-IDF

Show All Projects

Skills

Languages

JavaScriptTypeScriptPythonC++C

Frontend

React.jsNext.jsRedux ToolkitRTK QueryTanStack QueryVitePWA

Backend

Node.jsExpress.jsREST APIsWebSocketsJWTPassport.jsOAuth 2.0

Databases

MongoDBIndexing & AggregationQuery OptimizationMySQLSQL Server

Architecture

MicroservicesEvent-Driven SystemsMVCFactory PatternConcurrencyACID TransactionsRate LimitingIdempotent APIs

DevOps & Tools

GitBitbucket CI/CDJestAgile / Scrum

Contributions

“You don't get what you wish for, you get what you work for.”

— Virat Kohli

Let's work together

Have a project in mind? Let's create something amazing.

Email Me Book a Call