Hello, I am
Electrical & Electronics Engineering undergraduate at NIT Andhra Pradesh with a Minor in Quantum Computing and a CGPA of 9.4/10. My research focuses on Generative AI Safety & Privacy, machine learning, model evaluation, and reliable AI systems. I also build projects across computer vision, reinforcement learning, embedded AI, and intelligent software systems.
Lightweight safety firewall for obfuscated, multilingual and code-mixed prompts, combining normalization, multilingual risk detection, risk fusion and safe rewriting.
Hybrid CNN + BiLSTM + Attention framework for battery anomaly detection, SoC/SoH estimation, RUL prediction and EV range forecasting on 10,000+ records.
Autonomous companion robot integrating ESP32, Raspberry Pi, OpenCV and OpenAI API for sensor acquisition, navigation, motor control and natural interaction.
Full-stack AI platform that parses resumes and ranks candidates against job descriptions using NLP-based similarity scoring and recruiter dashboards.
Automated attendance system using dlib and CNN face recognition, achieving 92% accuracy across 30+ student profiles with real-time Google Sheets synchronization.
Custom MARL environment with decentralized reward structures, studying cooperative and competitive agent behavior across 8,000+ training episodes.
IISc Bangalore • Generative AI Safety & Privacy
Developing a multi-subject identity protection framework against DreamBooth-based personalization attacks in latent diffusion models. The work uses gradient-based attention attribution to identify identity-sensitive self-attention heads and construct shared identity subspaces for localized model immunization.
69%
Lower DINO score vs. APDM baseline
24×
Faster optimization time
0.137 → 0.042
DINO score improvement
IISc Bangalore | 2026
Research on diffusion-model personalization attacks, identity protection, attention attribution, localized immunization and quantitative safety evaluation.
NIT Andhra Pradesh | 2026–27
Leading technical research and publication initiatives covering research papers, emerging technologies, AI/ML trends and technical frameworks.
InAmigos Foundation | May–Jun 2026
Built data-analysis and AI-driven analytics pipelines to extract actionable insights from organizational datasets supporting campaigns.
Secured 4th Place among 85+ projects at NIT Andhra Pradesh Project Expo 2025.
Maintaining 9.4 CGPA at NIT Andhra Pradesh.
Won 3rd Prize at Coding Club Ideathon Competition.
© 2026 Poojari Nagashiva | AI Safety • Machine Learning • Research