Prince Kumar Prajapati
Software Developer & Systems Builder
Software Developer with a strong foundation in data structures, AI/ML integration, and multi-chain Web3 systems. Experience building end-to-end production platforms, including an AI-powered proctoring system for ONGC with sub-second grading, an agentic AI outreach engine built on Google ADK, and a cross-border Web3 payment platform on Stellar. Technical Head at GDG GGSIPU leading workshops for 800+ students. GATE 2026 qualified with an IEEE-published research paper.
- Architected and deployed an AI-powered proctoring platform using YOLOv8n, FastAPI, MongoDB (Motor), and WebSockets, cutting evaluation turnaround time from hours to under 1 second via an automated grading engine.
- Built a multimodal client-side computer vision and audio pipeline with MediaPipe FaceLandmarker and the Web Audio API, detecting 6+ cheating-behavior signals (head pose, iris gaze deviation, multi-face presence, lip movement, background talking) at sub-100ms latency.
- Designed an in-memory, discard-unless-flagged evaluation pipeline and a bidirectional WebSocket 3-strike anti-cheat engine, streaming live feeds to an Admin Command Center while cutting storage overhead by ~90%.
- Architected and delivered end-to-end, production-ready full-stack web applications for client projects, scaling clients' digital footprints.
- Engineered custom social automation tools and API integrations for Google and Meta Ads platforms, streamlining marketing workflows and improving campaign efficiency.
- Managed full-lifecycle deployment of client applications, including CI/CD pipeline setup and cloud infrastructure, achieving 99.9% uptime.
- Collaborated cross-functionally to translate client business requirements into robust technical specifications and scalable software.
- Spearheaded and architected a scalable, multi-chain smart contract auditing platform supporting Solana, Move, EVM, and non-EVM ecosystems for real-time automated vulnerability analysis.
- Achieved 87% audit accuracy and reduced average audit turnaround time by 95%, enabling near real-time security analysis for Web3 developers.
- Automated security checks and AST validations, making smart contract auditing 10x more affordable than traditional manual methods, and deployed the tool internally to accelerate dApp development.
- Architected clean React and Node.js REST services with containerized deployment using Docker.
- Engineered automated SEO pipelines that boosted client Google Search Index ranking by 80%.
- Streamlined state management and frontend reactivity using Redux Toolkit.
- Led 5+ major technical events and hands-on workshops engaging 800+ students in web development, AI/ML, and cloud computing.
- Organized campus hackathons, student mentoring tracks, and collaborative architecture sessions.
- Expanded community outreach 3x through strategic campaigns, increasing event participation and industry collaboration.
- Connected collegiate innovators with industry mentors, leading hackathons and developer outreach.
- Academic Standing: CGPA 92.53% (Awarded Certificate of Excellence for highest overall grades in College Department).
- Specialization Minor: Artificial Intelligence & Machine Learning (Minor: AI & ML).
- GATE 2026 Qualified in Computer Science and Information Technology.
- Active research author with publications in IEEE Xplore and DTC Journal of Computational Intelligence.
Qualified in Computer Science & Information Technology, demonstrating strong theoretical and applied mastery of algorithms, systems, and discrete mathematics.
Awarded Certificate of Excellence for achieving the highest overall grades (CGPA: 92.53%) in the Computer Science & Technology department.
Clinched top honors and prizes in 4+ hackathons across Web3, AI computer vision, and real-time proctoring systems.
Authored peer-reviewed clinical research on MindMedix: AI Platform for Neurological Diagnosis, published on IEEE Xplore.
End-to-end clinical diagnostic framework integrating Deep Convolutional Neural Networks (DCNN) and generative AI to detect and segment neurological pathologies on MRI scans. Authored with 4 researchers under the mentorship of Dr. Madhumita Mahapatra.
Presents an optimized Deep Convolutional Neural Network architecture tailored for real-time facial feature extraction, sentiment analysis, and multi-class emotion classification across edge computing devices.