Hey I'm Naman

I'm a |

Passionate About Transforming Data into Insightful Solutions

About Me

I am Naman V Srivastav, a recent Computer Science graduate specializing in AI and ML from PSIT, Kanpur. I’m passionate about developing impactful AI models to drive innovation in real-world applications. Throughout my academic journey, I engaged in rigorous research across areas like federated learning, quantum machine learning, and NLP, deepening my expertise and commitment to the field.

One of my key projects is a cardiovascular risk prediction system leveraging Federated Learning and Feedforward Neural Networks. This model, which achieved 92% accuracy, enhances patient engagement in cardiovascular health management by 30% while ensuring data privacy. My research on this work was published in the International Conference on Innovations in Computing (ICRIC-2023), Volume 2.

Additionally, I developed VIBE (Virtual Intelligence for Business Enhancement), an LSTM-based stock prediction model with 87% accuracy, and Project Opinify, a customer sentiment analysis tool for e-commerce. Opinify uses Flask for backend ML processing and Streamlit for a visually appealing, user-friendly interface, creating an accessible platform for actionable sentiment insights.

Achievements

Published my first research paper on "Heart Disease Prediction Using Federated Learning" in Springer, achieving 89.98% accuracy while ensuring privacy-preserving AI innovation.
Solved over 200+ complex problems on Leetcode, demonstrating expertise in algorithms and data structures.
Achieved top ranking and certification in Hackerrank algorithm challenges, demonstrating expertise in solving real-world problems.
Earned a Python Programming Certificate from Hackerrank, validating proficiency in Python through advanced coding challenges.
Participated in multiple hackathons like Flipkart 3.0, Hero Campus Challenge, Codeagoda, Flipkart 5.0, and Tata Steel TomorrowLab.

Research Experience

Heart Disease Prediction Using Federated Learning

Pioneered a privacy-focused cardiovascular risk prediction framework using FeedForward Neural Networks and Federated Learning, achieving 92% prediction accuracy while ensuring data security through decentralized model training.
Enhanced patient outcomes by increasing engagement in cardiovascular health management by 30% through a daily health tracking feature.
Earned recognition with the acceptance of a research paper by ICRIC-2023 (International Conference on Innovations in Computing), highlighting the innovation in predictive healthcare technology.

Leadership

PSIT R&D Department

Directed AI/ML research, developing algorithms that improved efficiency by 25% in Federated Learning, NLP, and CV.
Coordinated a team of 5 researchers, driving collaborative projects and facilitating knowledge sharing.
Spearheaded the organization of five seminars and workshops on cutting-edge AI technologies, attracting over 200 attendees; fostered a culture of continuous learning that increased team knowledge-sharing by 70% within six months.

Skills

Programming Languages

Data Analysis Tools

Frameworks & Libraries

Development Tools

Machine Learning & AI

Cloud Platforms

Other Tools

Programming & Development: Python, C/C++, SQL, Flask, Django, FastAPI, HTML/CSS, JavaScript

Machine Learning & AI: TensorFlow, PyTorch, Scikit-learn, Hugging Face, NLP, Federated Learning, XGBoost, OpenCV

Cloud & Data Management: AWS, Google Cloud, Docker, SQL, Vector Databases (Qdrant, Pinecone, FAISS)

Tools & Version Control: Git, GitHub, VS Code, Jupyter, Data Visualization (Matplotlib, Seaborn, Plotly)

Projects

Opinify

Technologies: NLP, Flask API

Integrated a pre-trained XGBoost model into a Flask API with 89% accuracy and developed a Streamlit frontend for intuitive sentiment analysis, achieving 95% user satisfaction.

Improved data quality through advanced preprocessing and implemented visualizations with Matplotlib, reducing analysis time by 30%.

GitHub Repository

V.I.B.E

Technologies: LSTM, RNN, Python

Virtual Intelligence for Business Enhancement

Designed advanced LSTM models for stock price prediction, achieving 87% precision and driving a 22% portfolio value growth in six months.

Leveraged RNNs for long-term dependency capture, reducing mean squared error by 15% for enhanced real-time market analysis.

GitHub Repository

AI Response Evaluator

Technologies: Python, WebSockets, Redis

Developed an AI-powered WebSocket-based response evaluation system using Sentence-Transformers for semantic similarity scoring, delivering real-time, context-aware feedback (e.g., ”Great answer!” for scores above 0.8).

Strengthened performance by integrating Redis caching to reduce evaluation time by 50% and ensured reliability through unit testing with Python’s UnitTest framework.

GitHub Repository

C.A.R.D.I.A

Technologies: FL, NN, Flower

Cardiovascular Risk Prediction using Distributed AI

Pioneered a privacy-focused cardiovascular risk prediction framework using FeedForward Neural Networks and Federated Learning, achieving 92% prediction accuracy while ensuring data security through decentralized model training.

Enhanced patient outcomes by increasing engagement in cardiovascular health management by 30% through a daily health tracking feature.

GitHub Repository

RAG Based ChatBot

Technologies: RAG, NLP, Vector DBs

Topic-Specific Chatbot with Retrieval-Augmented Generation (RAG)
Precision and Contextual Relevance: Combines Biomistral 7B for domain-specific language processing and Qdrant for efficient vector retrieval, ensuring accurate and relevant answers for healthcare and technical queries.
Optimized Query Processing: Utilized NeuML/pubmedbert-base-embeddings for semantic understanding and integrated advanced prompt engineering techniques such as few-shot learning and chain-of-thought prompting for nuanced responses.

GitHub Repository

Contact Me

Get in Touch

Contact Information

Email: vsnaman2002@gmail.com

Phone: +91 91561 79516

LinkedIn: https://www.linkedin.com/in/namanvsrivatav

GitHub: https://github.com/NamanVSrivastav