Passionate final year B.E. student specializing in Artificial Intelligence & Machine Learning. Experienced in building robust LLM systems, Generative AI pipelines, and custom Deep Learning architectures.
Reflecting strong technical rigor, high analytical aptitude, and consistent performance.
As a final year student specializing in Artificial Intelligence and Machine Learning, I am actively preparing for the future of enterprise automation and system orchestration.
My skill set matches professional engineering demands, specifically working with Retrieval-Augmented Generation (RAG) paradigms, model optimization, and contextual search systems. I am highly motivated to implement intelligent, high-accuracy products that streamline operations, with a strong focus on corporate data alignment, accuracy benchmarks, and scalability.
Developed a secure RAG-based AI chatbot using Streamlit, Azure OpenAI, OCR, embeddings, and vector databases to parse and analyze complex enterprise PDF documents.
Optimized search performance by mapping context-aware text chunks from database registers, feeding high-fidelity responses to the LLM for automated summarizing, predictions, and accurate metrics.
EEG-Based Emotion Recognition System using Brain Computing Interface (BCI)
Problem Statement:
Detecting human emotional states (happy/sad/neutral) in real-time from high-dimensional, noisy EEG signals.
Scientific Evaluation:
Evaluated performance of Neural Networks (NN), K-Nearest Neighbors (KNN), and Support Vector Machines (SVM) algorithms across 8 crucial brain regions (channels).
Benchmark Accuracy: Neural Networks consistently outperformed other baseline frameworks, yielding up to 99.8% precision rate on Target Channel 43.
Navigating, extracting, and querying massive legal corporate document agreements covering multiple years of wages.
Architectural Solution:
Architected an end-to-end RAG system utilizing Streamlit, Azure OpenAI (GPT-4o), OCR layers, and semantic embeddings for highly robust contextual answers.
System Predictions Validated: Successfully produced accurate financial predictions for HRA (6,500 - 7,000/-) and Basket Amount (15,000 - 16,000/-).