Druvith Gowda

I’m a AI Engineer exploring the frontiers of intelligent systems.

Ethos

I'm driven by a fundamental curiosity about intelligence—both human and artificial. My work is a continuous exploration of this, expressed through two passions:

  1. Building practical AI tools that enhance human capability.
  2. Researching the complex decision-making processes of autonomous agents.

For me, AI is not just a tool, but a new medium for creation and discovery. My goal is to build systems that learn, adapt, and solve problems in increasingly sophisticated ways.

Projects

Semantic Video Search

Ask questions and find exact moments in your video library. This system transforms video archives into a searchable knowledge base. By using multimodal AI to analyze and index video content, it allows users to find precise moments using natural language queries, eliminating the need for manual scrubbing.

Gemini Flash Pinecone React
Source

Self-Operating Computer

A framework that enables multimodal models to operate a computer, automating tasks by seeing the screen and using the mouse and keyboard. It could also be used as end-to-end web testing agent.

Python FastAPI Multimodal AI PyAutoGUI EasyOCR Docker
Source

MEDMISTRAL

Fine-tuned Mistral-7B model specialized in medical Q&A, diagnostic hints and treatment suggestions. Trained on 200 k high-quality medical dialogues.

Mistral-7B LoRA Medical AI
Hugging Face

Neural Networks from Scratch

Self-contained notebooks that build dense, CNN and RNN layers with pure NumPy. Includes math notes & visualizations for intuitive understanding.

Python NumPy ML Fundamentals
Source

Let’s Talk

If you reached this far and would like to know more, drop me a text!

druvithlgowda00@gmail.com