Sunny Kumar
AI EngineerAI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI Engineer with less than a year in Machine Learning & RAG systems
Sunny Kumar is an Integrated Dual Degree student in Computer Science & Engineering with AI, demonstrating strong skills in Machine Learning, Deep Learning, NLP, and RAG. With internship experience as an AI Intern at Prodigal AI and an AI/ML Intern at Infosys Springboard, Sunny has architected RAG-based systems, fine-tuned transformer models, and optimized NLP pipelines. He has also developed real-time commerce platforms and low-latency vision inference systems, showcasing expertise in full-stack development, distributed systems, and scalable inference solutions.
Rajiv Gandhi Institute of Petroleum Technology
Integrated Dual Degree · Computer Science & Engineering with AI
August 1, 2022 – June 30, 2027
Prodigal AI
AI Intern
May 1, 2025 – August 1, 2025
India
Infosys Springboard
AI/ML Intern
October 1, 2024 – December 1, 2024
India
Local Link: Real-Time Local Commerce Platform
January 1, 2023 – Present
Built a production-ready MERN commerce platform featuring separate customer and shopkeeper portals with secure JWT-based role authentication, inventory management, and online order processing. Implemented Socket.IO-based real-time order tracking with instant status updates, integrated Razorpay payments, and optimized dashboard performance using MongoDB compound indexing for low-latency queries.
View ProjectLow-Latency Vision Inference System
January 1, 2023 – Present
Built a production-grade client-server inference pipeline using FastAPI and YOLOv8; a stream-processing client captures frames from webcams, RTSP streams, or video files and sends them to a REST inference server returning structured JSON detections with latency metadata. Implemented an async queue with a thread-pool executor to decouple HTTP handling from GPU inference, achieving 50-100 FPS (10-20 ms/frame) while supporting concurrent multi-stream processing and exposing /health and /metrics endpoints.
View ProjectGovernment Schemes Info Retrieval System
January 1, 2023 – Present
Scraped 300+ government schemes and developed a full-stack AI-powered question answering platform using React, Node.js, MongoDB, and a RAG pipeline with LangChain and FAISS. Achieved 95% scraping success and 90% query accuracy, reducing manual lookup time by 70%; deployed on Vercel with secure OAuth 2.0 authentication.
View ProjectM.Tech Thesis: Dynamic Sparsity in Small Language Models
January 1, 2022 – Present
Investigating dynamic attention-head skipping techniques to improve inference efficiency in small language models. Developing adaptive sparsity mechanisms for Phi-4 Mini and Llama 3.2 to reduce memory usage and inference latency while maintaining model accuracy.
Cultural Fit Analysis
The candidate's diverse project portfolio, including full-stack development, AI/ML applications, and academic research, suggests adaptability and a broad interest in technology. Participation in hackathons and competitive programming indicates a collaborative and challenge-driven mindset. The projects align well with an AI Engineer role, demonstrating a strong fit for a technically demanding and innovative environment.
Soft Skills & Operational Fit
The candidate demonstrates strong problem-solving skills through competitive programming and hackathon wins. Project descriptions indicate an ability to work on complex, end-to-end systems and optimize for performance, suggesting a proactive and results-oriented work attitude. The focus on production-grade systems and real-time applications implies an understanding of operational requirements.