Also known online as: yashbytes
I build AI-powered products and the infrastructure they run on — an AI-first engineer shipping production systems, from LLMs to Linux.
Currently working on production AI features across live products. Final-year computer science student.
Currently exploring AI agent architectures, distributed systems, Linux internals, and networking.
- 🐧 Daily driving Arch Linux + Hyprland
- 🤖 Building AI systems: RAG pipelines, AI agents with tool use, LLM integrations
- 🌐 Exploring networking, DNS infrastructure, and self-hosting
- 🛠 Building with Python, TypeScript, FastAPI, Next.js, PostgreSQL, Docker, Bash, and C
- ⚡ Interested in AI engineering, platform engineering, and distributed systems
AI Engineering & LLM Systems
Software Development
Linux Internals
Networking & DNS
Infrastructure Automation
Containers & Deployment- 🤖 AI Bhai
PRIVATE— mobile-first consumer AI assistant. Chat Mode + task-oriented Work Mode, Hindi / Marathi / English + voice, MCP plugin ecosystem - 🔍 AtlasRAG — local-first enterprise RAG over PDF / TXT / Markdown with grounded answers and source citations
- 💬 ChatAPI — document-grounded AI agents for local businesses. Embeddable chat widget, JWT auth, dashboards
- 🔌 DesignMCP — MCP server giving AI coding agents UI inspiration. Query parsing, TTL cache, stdio transport
- 📝 Distributed T5 fine-tuning — built a compute cluster from university lab computers; fine-tuned T5-base for grammar correction
- 🐚 MYcShell — custom C shell for Unix + Windows. Systems programming, start to finish
- 🌐 my-own-internet — alt-root DNS infrastructure with custom TLDs. Real networking, not a toy
Software Development
AI Engineering & Agents
Linux
Networking & DNS
Infrastructure Engineering
DevOps & Platform Engineering
Distributed Systems
Systems Programming
Self-Hosting
Open Source- Website: https://yashpatil.dev
- LinkedIn: https://linkedin.com/in/yashbytes
- GitHub: https://github.com/Real-yash
- X: https://x.com/yashbytes
- Email: yashpatil.cse@gmail.com
I enjoy understanding how things work under the hood —
from LLM tool-calling loops to Linux kernels and networking stacks.