BARSHAN MAJUMDAR
@Barshan249
BARSHAN MAJUMDAR
@Barshan249
Machine Learning Engineer and Full Stack Developer
Machine Learning Engineer and Full Stack Developer
Kolkata, India
ā” Barshan Majumdar
Machine Learning Engineer | Full-Stack Developer
š Kolkata, West Bengal, India
āļø [email protected]
š Portfolio: https://barshan-majumdar.vercel.app
š GitHub: https://github.com/Barshan-Majumdar
š LinkedIn: https://linkedin.com/in/barshan-majumdar
š About Me
Machine Learning Engineer and Full-Stack Developer who ships resilient, scalable systems when the clock is ticking. I specialize in multi-modal AI pipelines, local LLM orchestration, and robust web architectures rather than just API wrappers. With a published research chapter (Springer Nature) and a consistent track record of podium finishes at competitive hackathons, I build production-ready applications that solve real-world problems.
š Hackathon Track Record
- š„ Winner - Signifiya (Adamas University)
- š 4th Place - HackStorm 2.26, BPPIMT (ShieldX)
- š Top 20 - HackOdisha, 2025 (Learn Flow)
- š» Finalist / Participant - Google Solution Challenge 2026 & SIH 2025
š» Tech Stack
- Languages: TypeScript, JavaScript (ES6+), Python, SQL, C/C++
- Frontend: Next.js, React, Tailwind CSS, Framer Motion
- Backend: Node.js, Express.js, NestJS, REST APIs, JWT Auth
- Databases: PostgreSQL, MongoDB, Redis
- AI & ML: Local LLM deployment (Llama, Gemma), Hugging Face, Multi-modal semantic validation (CLIP), Model fine-tuning
- DevOps & Tools: Docker, AWS, Vercel, GitHub Actions, CI/CD, Git
š ļø Featured Projects
1ļøā£ SevaSetu
TypeScript, Next.js, CLIP Model, PostgreSQL, Node.js
- Live: https://sevasetu.vercel.app
- Source: https://github.com/Barshan00/SevaSetu
- Engineered a humanitarian coordination and crisis response platform submitted for the Google Solution Challenge 2026.
- Integrated a custom CLIP-based semantic verification model to authenticate disaster relief requests and filter anomalies.
- Implemented real-time geospatial routing and role-based workflows for emergency responders.
2ļøā£ ShieldX
TypeScript, Next.js, Python, OpenCV, WebSockets
- Live: https://shieldx-proctor.vercel.app
- Source: https://github.com/Barshan00/ShieldX
- AI-powered exam proctoring and integrity monitoring system built during HackStorm 2.26.
- Powered by a multi-modal vision engine for real-time anomaly detection, head-pose tracking, and automated logging.
- Utilized low-latency WebSocket streaming for instant violation alerts to administrators.
3ļøā£ Learn Flow
TypeScript, Next.js, Gemini API, MERN Stack
- Live: https://learn-flow.vercel.app
- Source: https://github.com/Barshan00/Learn-Flow
- AI-powered study planner that parses PDF syllabi and generates structured, chapter-wise learning schedules.
- Achieved a Top 20 ranking nationally in December 2025 alongside team Code Hustlers.
- Built custom breakdown logic powered by Gemini to tailor study modules to individual pacing.
4ļøā£ AeroGuard
Next, Tailwind, Framer Motion, FastAPI, Ultralytcs, Neon, Drizzle, Cloudflare R2, Nginx
- Live: https://aero-guard-plane.vercel.app/
- Source: https://github.com/Barshan-Majumdar/AeroGuard.git
- AeroGuard is an enterprise-grade, intelligent aircraft inspection platform. It replaces manual, error-prone borescope video reviews with automated defect detection, interactive 3D digital twin visualization, and compliance-ready reporting. (AeroGuard is an enterprise-grade, intelligent aircraft inspection platform. It replaces manual, error-prone borescope video reviews with automated defect detection, interactive 3D digital twin visualization, and compliance-ready reporting.
Experience
Elenxia
Campus Ambassador & Hero Head (Remote, Aug 2025 - Sep 2025)
-
Promoted skill-based learning platforms to 200+ students.
-
Organized and led 5+ hackathons and technical virtual workshops.
LetsUpgrade
Student Ambassador (Remote, June 2025 - July 2025)
Open Source
- GSSoC Contributor - 2025
š Research & Publications
- Springer Nature: Published a peer-reviewed research paper chapter integrating advanced machine learning methodologies into practical digital workflows. (https://link.springer.com/chapter/10.1007/978-3-032-16711-8_2)
š Education
- Brainware University | B.Tech in Computer Science and Engineering (AI & ML) | July 2024 - August 2028
šÆ CGPA: 9.65