Anubhav Sanket

I'm a Machine Learning Engineer based in Bangalore who loves building AI solutions that actually work in the real world. My focus? Offline-first systems that respect privacy and run fast without any cloud dependency, no data leaks, just efficient AI that gets the job done.

Recently, I built Viral Clipper, a local AI video platform using Gemma 3 and WhisperX that cuts viral moments from long videos completely offline. It replaces $360/year cloud tools with zero recurring costs. At Oaz Shakti, I created dashboards tracking ₹124.6M in transactions and flagged ₹8.4M worth of inconsistent payments, helping the business manage cash flow better. Before that, I shipped an e-commerce site at Varcons that loaded 30% faster.

I'm competitive but practical. I placed Top 3.5% globally (158/4,506) in a Kaggle finance competition, published research on customer segmentation at IJNRD, and participated in Google's Tunix Hackathon. Outside tech, I organized cultural events with the Art & Literature Club, turns out I enjoy bringing people together as much as I enjoy optimizing code.

I hold a B.E. in AI/ML from Dayananda Sagar Academy (2024) and picked up certifications in Azure Data Engineering, Power BI, and Cyber Security along the way.

Right now, I'm looking for ML Engineer / GenAI Engineer roles where I can build things that matter, whether that's deploying local LLMs, speeding up video processing with CUDA, or creating analytics that drive real decisions. I'm especially interested in roles that involve:
- Building and deploying LLMs (local or cloud)
- Multimodal AI (audio, video, text)
- Privacy-preserving AI systems
- Performance optimization (CUDA, quantization, etc.)
- Real-world AI applications that solve actual business problems