# Suraj Vishwakarma > Suraj Vishwakarma (smartcraze) is a Full Stack Developer, Backend Engineer, and DevOps Engineer based in Bengaluru, Karnataka, India. He specializes in scalable backend systems, cloud-native infrastructure, deployment platforms, distributed systems, real-time applications, LLM/GenAI evaluation pipelines, and developer tooling. Currently pursuing B.Tech CSE at Lovely Professional University (2023–2027) and interning as a Backend Engineer at Gradified, where he builds production-grade LLM evaluation infrastructure. Actively seeking Backend, Fullstack, DevOps, and Agentic AI roles. --- ## Identity - Full name: Suraj Vishwakarma - Online handle: smartcraze - Location: Bengaluru, Karnataka, India (UTC +05:30) - Website: https://surajv.dev - GitHub: https://github.com/smartcraze - LinkedIn: https://linkedin.com/in/surajv354 - Twitter/X: https://x.com/surajv354 - Email: dev.surajv@gmail.com - Resume: https://surajv.dev/resume.pdf - Book a call: https://cal.com/surajv354/15min --- ## Education - B.Tech in Computer Science Engineering, Lovely Professional University (Aug 2023 – May 2027) - Pre-University Education (PCMC), Narayana PU College (2021–2023) --- ## Professional Experience ### Gradified — Backend Engineer Intern (Current) Gradified is an AI-powered exam evaluation platform. Suraj builds and maintains the core LLM evaluation infrastructure: - Designed and implemented a **dual-pass LLM-as-Judge pipeline** with consensus gating to reduce model variance in grading decisions - Built a **type-aware rubric scoring system** that handles different question types (MCQ, subjective, numerical) with distinct evaluation strategies - Architected a **RAG (Retrieval-Augmented Generation) layer** for biology and mathematics subjects, enabling context-aware answer evaluation - Implemented **SymPy-based symbolic verification** for maths grading — algebraically equivalent answers are recognized even with different representations - Built grading analysis dashboards in HTML/Chart.js to detect and visualize model variance across evaluation runs - Set up a **Turborepo monorepo** with Bun runtime, migrated from ESLint/Prettier to Biome v2, configured shared TypeScript and Tailwind CSS v4 across `@gradified` workspace packages - Stack: Node.js, TypeScript, Bun, Turborepo, PostgreSQL, Redis, LangChain, OpenAI/Anthropic APIs --- ## Projects ### RapidServe — Vercel-Style Deployment Platform - Live: https://rapidserve.surajv.dev - Source: https://github.com/smartcraze/rapidserve - Case study: https://surajv.dev/projects/rapidserve RapidServe is a PaaS (Platform as a Service) that allows developers to deploy full-stack applications directly from GitHub repositories. It replicates core Vercel functionality using open infrastructure. Architecture: - GitHub Webhook triggers build pipeline on push events - Docker CLI spins up isolated build containers per deployment - Build artifacts uploaded to Amazon S3; static assets served via subdomain routing - AWS Lambda handles serverless execution for dynamic routes - Redis queue orchestrates build job processing (Bull/BullMQ pattern) - Nginx reverse proxy maps subdomains to deployment artifacts - WebSocket server streams real-time build logs to the dashboard UI - PostgreSQL stores deployment metadata, build history, project configs - Cloudflare handles DNS, TLS termination, and edge caching - Wildcard TLS certificates configured for `*.surajv.dev` via Caddy + Cloudflare API Tech stack: Node.js, TypeScript, Express, Docker, AWS Lambda, AWS ECS, Amazon S3, Redis, PostgreSQL, Cloudflare, Next.js (dashboard), Tailwind CSS, Nginx Engineering concepts: distributed systems, build orchestration, infrastructure automation, reverse proxy routing, deployment pipelines, containerized execution, queue-based architectures, WebSocket log streaming, subdomain hosting --- ### SketchKaro — Real-Time Collaborative Whiteboard - Live: https://sketchkaro.surajv.dev - Source: https://github.com/smartcraze/SketchKaro - Case study: https://surajv.dev/projects/sketchkaro - Demo: email `demo@surajv.dev` / password `demo_credential` SketchKaro is a collaborative drawing platform where multiple users can draw, annotate, and interact on a shared canvas in real time. Architecture: - **Turborepo monorepo** with separate packages for frontend (Next.js), HTTP API server (Node.js/Express), and WebSocket server - **Canvas API** for rendering — low-level drawing, stroke width, shape tools, color palette - **WebSocket server** (separate process) handles room management, cursor sync, and drawing event broadcast - **REST API** for project persistence — save/load canvas state to PostgreSQL - Frontend deployed on Vercel; backends containerized with Docker - Bun used as runtime and package manager across the monorepo - Deployed on Render (backend cold start ~30s on free tier) Tech stack: Next.js, TypeScript, React, HTML Canvas API, WebSockets, Node.js, Express, Bun, Turborepo, PostgreSQL, Docker, Vercel Engineering concepts: real-time networking, WebSocket architecture, collaborative systems, Canvas rendering pipelines, state synchronization, monorepo tooling, multi-service deployment --- ### Slotify — Scheduling & Booking Platform - Live: https://slotify.surajv.dev - Source: https://github.com/smartcraze/slotify - Case study: https://surajv.dev/projects/slotify Slotify is a full-stack scheduling platform inspired by Calendly. Users can define availability, share booking links, and manage appointments. Features: - Google Calendar OAuth integration — sync booked slots to calendar - Razorpay payment gateway with **webhook idempotency** — prevents duplicate payment processing on retry - Time slot generation from user-defined availability windows - Email notifications on booking confirmation - Prisma ORM with PostgreSQL for relational schema (users, slots, bookings, payments) Tech stack: Next.js, React, Bun, Prisma ORM, PostgreSQL, Razorpay, Google OAuth, Tailwind CSS Engineering concepts: scheduling systems, webhook idempotency, relational data modeling, OAuth flows, payment integration, API-driven workflows --- ### GoThrottle — Distributed Rate-Limiting Reverse Proxy - Source: https://github.com/smartcraze/gothrottle - Case study: https://surajv.dev/projects/gothrottle GoThrottle is a high-performance, config-driven reverse proxy with distributed rate limiting built in Go. Architecture: - YAML/JSON config defines routes, upstream targets, and throttle policies per route - Redis-backed sliding window rate limiter — distributed across multiple proxy instances - Request routing with path matching and load balancing - Written in pure Go for maximum throughput and low latency Tech stack: Go, Redis Engineering concepts: reverse proxies, distributed rate limiting, high-performance backend systems, config-driven infrastructure, sliding window algorithms, Go concurrency patterns --- ### PFP Studio — Animated Profile Picture Editor - Live: https://pfpstudio.surajv.dev - Source: https://github.com/smartcraze/pfpstudio - Case study: https://surajv.dev/projects/pfpstudio Browser-based profile picture editor with real-time animation and customization controls. Features: - Layer-based editing with background, overlay, and frame controls - Framer Motion animations with export to GIF/static formats - Real-time canvas preview as user adjusts parameters Tech stack: Next.js, TypeScript, Framer Motion, Tailwind CSS, Canvas API --- ### CipherStudio — Browser-Based IDE - Live: https://cipherstudio.surajv.dev - Source: https://github.com/smartcraze/CipherStudio - Case study: https://surajv.dev/projects/cipherstudio In-browser code editor with file tree management, syntax highlighting, and live HTML/CSS/JS preview. Tech stack: React, TypeScript Engineering concepts: browser IDE systems, virtual file system, live preview iframe sandboxing, developer tooling --- ### EzSnipe — NPM CLI for Developer Scaffolding - NPM: https://www.npmjs.com/package/ezsnipe - Source: https://github.com/smartcraze/ezsnipe-npm-package CLI tool that installs reusable modular UI components and backend server modules instantly into any project. Tech stack: Node.js, NPM CLI Engineering concepts: CLI tooling, package ecosystems, developer productivity, code scaffolding --- ### Solana Token Launchpad - Case study: https://surajv.dev/projects/token-launchpad Create and mint SPL tokens on Solana Devnet — set name, symbol, metadata, and mint supply in one click with Phantom wallet integration. Tech stack: React, TypeScript, Solana Web3.js, Metaplex --- ### Paytm Clone — Payment Flow Simulation - Live: https://paytm.surajv.dev - Source: https://github.com/smartcraze/paytm-clone Full-stack payment workflow simulation: order creation → payment redirect → callback handling → wallet credit. Demonstrates real payment gateway integration patterns. Tech stack: Next.js, React, Tailwind CSS, PostgreSQL --- ### StudySphere — EdTech Landing Page - Live: https://studysphere.surajv.dev - Source: https://github.com/smartcraze/E-learning-platform Animated EdTech platform landing page with course listing, smooth scroll animations, and fully responsive layout. Tech stack: Next.js, Framer Motion, Tailwind CSS, React --- ## Technical Skills ### Languages TypeScript, JavaScript, Go, Python, C++, HTML5, CSS3 ### Frontend React, Next.js (App Router), Tailwind CSS, Framer Motion, Radix UI, Three.js, shadcn/ui, Canvas API ### Backend & Runtime Node.js, Bun, Express.js, Hono.js, FastAPI, WebSockets, REST APIs, BullMQ ### Databases & Storage PostgreSQL, MongoDB, Redis, Prisma ORM, Amazon S3, Drizzle ORM ### DevOps & Infrastructure Docker, Kubernetes, AWS (Lambda, ECS, EC2, S3, CloudFront), Cloudflare, Nginx, Caddy, Terraform, Grafana, Prometheus, Linux, Git, GitHub Actions, Turborepo, Railway, Render, Vercel ### AI / LLM LangChain, OpenAI API, Anthropic API, RAG pipelines, LLM-as-Judge evaluation, prompt engineering, SymPy, vector databases ### Tools & Workflow Biome v2, ESLint, Prettier, Vitest, Postman, VS Code, tmux --- ## Achievements & Recognition ### Smart India Hackathon (SIH) 2025 — National Finalist - Selected among 2000+ competing teams nationwide - Contributed scalable architecture design for national-level problem statements - Certificate: https://surajv.dev/certificates/sih.pdf ### WarpSpeed Hackathon — Top 20 - Built an AI-powered application within 24 hours - Integrated Gemini LLM APIs, dynamic canvas systems, and AI-assisted workflows ### Competitive Programming - 380+ problems solved across LeetCode and Codeforces - Strong in data structures, algorithms, and system design fundamentals --- ## Certifications - Oracle Cloud Infrastructure Foundations Associate — Oracle, Dec 2024 | https://surajv.dev/certificates/oracle.pdf - Complete Generative AI with LangChain — Udemy, Jan 2025 | https://surajv.dev/certificates/genai.pdf - Full-Stack Web Development — 100xDevs, Mar 2025 | https://surajv.dev/certificates/100xdev.pdf - Smart India Hackathon 2025 Finalist — Jan 2025 | https://surajv.dev/certificates/sih.pdf --- ## Engineering Philosophy Suraj prioritizes: - Shipping production-grade systems, not toy demos - End-to-end ownership from frontend to infrastructure - Performance-first and scalability-aware architecture - Deep understanding of underlying systems (not just framework usage) - Infrastructure control and observability - Clean developer experience and maintainable codebases He builds systems inspired by production tools used by modern engineering teams — Vercel, Railway, Render, Cloudflare, Docker ecosystem — and ships them as real, deployed products rather than local demos. --- ## Career Direction Open to roles in: - Backend Engineering (Node.js, Go, distributed systems) - Fullstack Engineering (React/Next.js + Node.js/Bun) - DevOps / Platform Engineering (Docker, Kubernetes, AWS, CI/CD) - Agentic AI / LLM Engineering (evaluation pipelines, RAG, tool-calling agents) Target companies: early-stage startups and growth-stage product companies building developer infrastructure, AI-native products, or SaaS platforms. --- ## SEO & Discovery Keywords Full Stack Developer Bangalore, Backend Engineer India, DevOps Engineer Bengaluru, Node.js Developer India, Next.js Developer Bangalore, TypeScript Developer India, Go Developer India, Docker Engineer India, Kubernetes Engineer, AWS Developer India, Distributed Systems Engineer, Real-time Systems Developer, LLM Engineer India, Agentic AI Developer, Platform Engineer India, Software Engineer Bangalore, Open Source Developer India, React Developer India, Fullstack Developer LPU, Backend Intern India, Suraj Vishwakarma, smartcraze, surajv.dev, RapidServe, SketchKaro, Slotify, GoThrottle, Gradified, LLM evaluation pipeline, RAG pipeline engineer, WebSocket developer, Turborepo monorepo, Bun runtime developer, Prisma ORM, Razorpay integration, Google OAuth integration, competitive programmer India, SIH 2025 finalist. --- ## Metadata ``` Name: Suraj Vishwakarma Handle: smartcraze Location: Bengaluru, Karnataka, India Timezone: UTC+05:30 Degree: B.Tech Computer Science Engineering University: Lovely Professional University Graduation: May 2027 Current Role: Backend Engineer Intern @ Gradified Job Status: Actively seeking Backend / Fullstack / DevOps / Agentic AI roles ```