I’m an AI Engineer who focuses on making models actually usable in the real world — deployed, optimized, and running reliably.

• 🎓 3rd-year Computer Science Engineering student.
• 🧠 Work spans LLMs, RAG systems, agent-based architectures, and local model inference
• ⚙️ Strong focus on deployment, servers, tunnels, automation, and infra
• 🔐 Interested in privacy-first AI (local & decentralized models)
• 🧪 Prefer building end-to-end systems over isolated demos

I care less about model hype and more about systems that work under constraints.

• Local LLMs (Ollama, Phi, Qwen, Gemma)
• RAG pipelines (CSV / vector DB / symptom-based retrieval)
• Tokenization & dataset preprocessing at scale
• Agent-based & multi-agent reasoning systems

• FastAPI-based AI services
• API orchestration & fallback strategies
• Real-time transcription & processing pipelines

• Cloudflare Tunnels (no open ports, ISP-safe)
• Nginx reverse proxy setups
• Self-hosted AI servers (CPU-constrained environments)
• GitHub Pages, Render, Streamlit deployments

• 🧠 Micropulse – Local & distributed SLM system (privacy-first AI)
• 🧩 RAG-based ERM Assistant – Reduced API calls by storing symptom-level abstractions
• 🎙 Live Audio Transcription + Keyword Extraction for medical workflows
• 🌍 Location-based Chat App – Geo-matching + WebSocket privacy model
• 🤖 Multi-agent Deliberative AI System (Project AETHER)
• 🧪 Trained & optimized small language models under limited compute
• 📄 Published 2 AI/ML research papers
• 🏆 Participated in multiple ML hackathons
• 💼 Internships: Doctors App, Mindray

If you care about AI that actually runs, not just models that look good in notebooks — we’ll get along.

📩 yashas.vaddi12@gmail.com 🔗 linkedin.com/in/yashasvaddi 🐙 github.com/YashasVaddi

“Build systems, not demos. Optimize for reality.”