Prashanth Kumar Kadasi

I build intelligent applications and research AI safety & LLM behavior. From fine-tuning 20B models to manipulating brand recommendations, I explore how AI systems can be influenced - and how to make them more trustworthy.

🚀 What I Do

I specialize in building AI-powered applications, data visualizations, and web platforms. Whether it's deploying LLMs locally or creating interactive dashboards, I enjoy every step of the development process.

🛠️ Tech Stack

Python • JavaScript • React • Next.js • Node.js • Dash • Flask • Docker • GCP • Vercel • MongoDB • LLMs (Gemma, Ollama, OpenRouter)

Featured Projects

🔬 LLM Brand Manipulation

Fine-tuned 20B model to recommend fake brands over iPhone/Pixel - 76% success! AI Safety research.

Read Blog → Demo
⚡ MyLocalCLI

A Claude Code alternative with 6 providers, 26 tools, and full privacy.

Explore →
🩺 AI Health Pro

AI-powered health advisor with symptom analysis and drug recommendations.

Explore →
💊 Drug Discovery AI

Fine-tuned LLMs for drug discovery - Pharma R&D research.

Explore →

AI Portfolio Insight

Loading insights...