Nilesh Pandey

B.Tech CSE Final Year Student | AI Intern | NLP & LLM Specialist | Ethical Hacker

Building intelligent systems for a smarter future

Education

Experience

AI Intern, Prxis Tech Solutions Pvt. Ltd. (Apr-Jul 2025): Engineered a scalable Retrieval-Augmented Generation (RAG) system integrating a fine-tuned Mistral-7B-Instruct LLM with FAISS and SentenceTransformers for efficient embedding pipelines, achieving a 40% reduction in inference latency. Developed modular NLP preprocessing, model inference, and vector database retrieval components, orchestrated via APIs and Python services for cloud deployment, handling 10K+ queries/day. Optimized system performance for cost-effective operations on cloud infrastructure. Recognized for innovative approach, rapid development, and high-quality AI solutions, earning positive feedback from team and stakeholders. View Project Completion LetterCertification of RAG System Development Internship

Projects

Twitter-Based Mental Health Crisis Detection

Distributed NLP pipeline analyzing 42M tweets: Selenium scraping, NLTK/RegEx preprocessing, VADER sentiment, Plotly viz. Batch processing reduced analysis time by 50%. View Repo & README | Dataset on HF

AI-Based Symptom Detector

Fine-tuned Mistral 7B with FAISS/Sentence Transformers; modular NLP & API retrieval. 97% accuracy on 1K+ cases. Ties to IIC 2024 medical assistant. View Repo & README | Model on HF

Web-Based Facial Authentication with Anti-Spoofing

TensorFlow/OpenCV/MediaPipe system replacing CAPTCHAs; custom CNN for anti-spoofing, SSL security. 98% accuracy, 85% spoofing drop. (Repo private – contact for details)

Achievements

Grand Finalist & 1st in Healthcare, IIC 2024: Top 10/764 teams; 48-hour AI medical assistant with Mistral 7B RAG for symptom diagnosis.

Technical Skills

Python
ML/AI (TensorFlow, PyTorch)
NLP/LLMs

Certifications