$ cat about.mdStudent : Odisha University of Technology and Research (OUTR), Bhubaneswar
Focus : Deep Learning · Agentic AI · AI Architectures
Building : RAG-BASE-AI-TEACHING-ASSISTANT
Learning : Advanced DL · Vector Databases · RAG & LLM Integrations
Goal : AI Engineer — blending creativity and code
$ tree stack/stack/
├── languages/ python · java
├── frameworks/ flask · fastapi
├── ai-ml/ langchain · llamaindex · openai · ollama · huggingface
├── vector-db/ qdrant · chromadb
├── ml/ scikit-learn · tensorflow
├── databases/ sqlite · mysql
└── tools/ git · github · linux · docker
$ cat agentic-ai.md> Designing agents that reason, plan, and act — not just autocomplete.
reasoning multi-step task decomposition, self-correction loops
memory RAG pipelines over vector stores for long-term context
tool-use function-calling, structured JSON output, API routing
orchestration multi-agent handoff and pipeline coordination
$ ls projects/ --sort=recentRAG-BASE-AI-TEACHING-ASSISTANT
RAG teaching assistant for Krish Naik's NLP (Hindi) course. Python · Ollama · Flask
structured-api-router
Fine-tuned Llama-3 (QLoRA) pipeline converting unstructured text into validated JSON API payloads. 100% schema compliance
data-science End-to-end Data Science journey — Python, ML, DL, NLP, deployment.
COMPLETE-DSA-LEARNING Daily DSA practice, pushed consistently.
