- 🔭 10+ years in IT — 7 yrs Backend Java, 4 yrs deep in GenAI
- 🤖 Building Pega Autopilot — LLM-driven AI assistant at enterprise scale
- 🎓 M.Tech AI/ML @ BITS Pilani (WILP) — 2025–2027
- 🤗 Published models on Hugging Face (420+ downloads)
- 🏢 Previously at Ericsson (SoftBank IoT), Mercedes-Benz R&D, CrimsonLogic
- 🇮🇳 Based in Bangalore, India
- 🐍 Member of PyTorch Day India HF community
🤖 LLM Fine-tuning ████████████████████░ 93%
🔗 RAG Systems ████████████████████░ 91%
🧩 Agentic / Multi-Agent ███████████████████░░ 90%
✍️ Prompt Engineering ████████████████████░ 92%
☕ Java / Spring Boot █████████████████████ 94%
🐍 Python / FastAPI ████████████████████░ 91%
| Project | Description | |
|---|---|---|
| 🧠 | qwen3-eng-odia-translation | Fine-tuned Qwen3 0.8B for English→Odia translation · 420+ downloads |
| 🌍 | eng-odia-translation | Live Gradio Space — try the model instantly |
| Repo | What it is |
|---|---|
| LLMs-from-scratch | Build a ChatGPT-like LLM in PyTorch from scratch |
| DeepSeek_From_Scratch | DeepSeek architecture — MoE, attention, modern LLM internals |
| all-agentic-architectures | 17+ agentic patterns — ReAct, tool-use, multi-agent orchestration |
| advanced_finetuning_bootcamp | LoRA, QLoRA, full fine-tuning techniques for LLMs |
| kafka_mcp_server | MCP server for Apache Kafka — AI agents + Kafka infra |
| rag_sessions | RAG techniques — naive, advanced retrieval, re-ranking, hybrid |
GenAI & LLMs
Backend & Data
DevOps & Cloud
AI Evals for Everyone · OpenClaw Mastery · RL for LLMs · AI Agent with N8N · Modern AI Pro (Practitioner) · Modern AI Pro (Essentials) · NLP Classification & Vectors · TensorFlow for AI/ML/DL · GenAI with LLMs · Python Programming · Java Generics · Pega CSA
💡 I build intelligent systems that decide, reason, and act.
From LLM-powered autopilots and multi-agent architectures to enterprise-grade NLP platforms.