I focus on building production-grade software that performs reliably at scale. Every engineering claim here is backed by system design schematics, live code walkthroughs, and automated load tests.

Secure online assessment platform built from scratch and deployed college-wide at KITS. Architected, built, deployed, and maintained entirely by me.

• Load test metrics (20-minute peak test): Successfully processed 146,045 total HTTP requests at a steady 118.7 req/s throughput with 100.00% request success (0 errors) and a P99 latency of 17.12 ms.
• Performance Engineering: Enabled a 3-tier leaderboard cache (local LRU cache with stale-while-revalidate ➔ shared Redis ➔ PostgreSQL with request deduplication and query coalescing) to prevent database stampedes under concurrent submissions.
• Memory Optimization: Fixed worker cache eviction storms by tuning memory pressure middleware limits (raising threshold to 1200MB) and resolving garbage collection event-loop blocking.
• Security & Anti-Cheat: Implemented full-browser protection including active tab-switch tracking, keyboard event interception, and DevTools/AI extensions blocking.
• CI/CD Pipeline: Configured automated GitHub Actions workflows running compilation checks, migrations, Jest tests, and Playwright E2E browser tests on spin-up PostgreSQL/Redis containers.

AI-powered Patient Acquisition & Trust Engine for dental restoration clinics. Built from scratch in a 12-hour sprint across 3 live evaluation phases evaluated by industry experts.

• Interactive Smile Visualizer: Developed interactive modules to project post-treatment dental restoration previews.
• Dental Tourism Calculator: Built comparison engines to estimate dental costs across multiple states and countries.
• Smart Follow-up System: Created an AI-driven automated message system to guide post-treatment patient recovery.
• Testing Integrity: Setup comprehensive unit and integration testing pipelines to verify patient cost calculation logic.

Production-grade platform replacing 100% of manual placement department paperwork at KITS, currently serving the administration, students, and recruiters.

• Administrative Dashboard: Real-time placement analytics and student registration tracking, supporting 300+ concurrent active users during placement drives.
• Student Portfolios: Manages individual academic scores, placement histories, and training progress with role-based access control.
• Reporting Automation: Automated Excel/PDF report generation, reducing administrative reporting time by ~80%.

Full-stack AI interview simulation platform with local LLM inference, multi-round placement workflows, and recruiter portfolio sharing — built for complete on-premise data privacy.

• Local LLM Inference Pipeline: Runs Qwen 2.5 3B via Ollama for real-time answer evaluation and dynamic question generation — zero external API dependencies, zero data leaks.
• Multi-Round Simulation Engine: Orchestrates Aptitude → Technical → HR interview rounds with configurable score-gating thresholds. Candidates must pass each round to advance, mirroring real placement drives across 23 companies (TCS, Google, Amazon, Goldman Sachs, etc.).
• Evaluation Queue with Concurrency Control: Priority-based task queue with bounded parallelism (max 2 active LLM calls), retry backoff (3 attempts × exponential delay), and heuristic load shedding at 100 pending tasks.
• Recruiter Portfolio Sharing: Students can generate privacy-masked public report links for recruiters — candidate emails, roll numbers, and voice/video logs are stripped from the public REST API response.
• 25,000-Sample Fine-Tuning Dataset: Generated a structured JSONL training corpus covering 2025-26 trending topics (RAG, Vector Databases, LLM fine-tuning, System Design) for offline model improvement.
• Admin Campaign Scheduler: Placement officers can schedule automated mock interview drives targeting specific departments, with a background worker that auto-enrolls students at the configured launch time.

💼  Full-Stack Developer Intern — Cognifyz Technologies                    Jun – Jul 2025
    React · Node.js · Express · MongoDB
    ➔ Developed RESTful API endpoints and integrated UI screens
    ➔ Optimized backend controller logic, reducing response latency by 30%

💼  Frontend Developer Intern — Codegnan (Offline)                          May – Jun 2025
    React · Figma
    ➔ Designed responsive UI prototypes and built modern frontend screens
    ➔ Optimized asset sizing to maintain quick landing page load times (<2s)

🥇  1st Prize — Codegnan Elevate X Hackathon                                May 2026
    Global Smile  ·  12-Hour Sprint  ·  React · Node.js · Prisma · PostgreSQL
    ➔ Built trust and billing engines for dental clinics under extreme time constraints
    ➔ Evaluated by technical panel across 3 development iterations

🥉  3rd Prize — Department Hackathon                                        Feb 2026
    StudentConnect  ·  48-Hour Sprint  ·  Python · Django · PostgreSQL
    ➔ Engineered resource-sharing platforms for academic departments

🌐  Google Student Ambassador                                              Jul – Dec 2025
    KITS Akshar Institute of Technology
    ➔ Organised 7+ developer workshops covering Android, cloud setups, and AI tools
    ➔ Built and managed a tech community of 300+ active student developers
    ➔ Drove a 50% increase in developer student registrations on campus

🛠️ Also Certified: ServiceNow Certified System Administrator (CSA) · CAD in progress — enterprise ITSM track