Final-year B.Tech (AI/ML) student focused on building real backend AI systems — retrieval that's accurate, APIs that are testable, code structured for actual use.
Proved this during my internship at ALIMCO, a Government of India PSU, building an enterprise RAG chatbot end-to-end.
Background: Diploma in Electronics Engineering (SBTE Bihar, 2022–2025) → B.Tech CSE (AI/ML), CGC Jhanjeri.
- 🔧 Refining backend AI systems and RAG pipelines
- 📚 Deepening LLM engineering and prompt design
- 🧩 Exploring agentic AI workflows
- 🎓 Completing my B.Tech (AI/ML) — final year
Artificial Intelligence Machine Learning Generative AI RAG Systems
LLMs Prompt Engineering FastAPI Backends
AI Intern — ALIMCO Artificial Limbs Manufacturing Corporation of India · Government of India PSU
- Built an enterprise RAG chatbot for semantic search & Q&A over assistive-device documentation
- Designed a modular FastAPI backend — loader → chunker → embedder → vector store → retriever → prompt → LLM
- Implemented semantic search with Sentence Transformers (BAAI/bge-base-en-v1.5) + ChromaDB
- Integrated dual LLM paths: Gemini API (cloud) and Ollama (local)
- Tested end-to-end via FastAPI Swagger UI — evaluated Easy 20/20, Medium 27/30, Hard 9/10, Expert 9/10
Enterprise RAG chatbot enabling semantic search and Q&A over assistive-device documentation. Backend-only, tested end-to-end via Swagger UI.
Document → Chunking → Embeddings (Sentence Transformers)
→ ChromaDB → Retriever → Prompt → LLM (Gemini / Ollama)
→ Answer
Tech: Python FastAPI Sentence Transformers ChromaDB Gemini API Ollama
Enterprise-style ML system predicting customer churn, deployed as an API with interactive dashboards.
Tech: Python Flask XGBoost Machine Learning
Customer support chatbot combining ML-based intent handling with a FastAPI backend, Streamlit frontend, and MongoDB storage.
Tech: Python FastAPI Streamlit MongoDB
| Languages |
|
| AI / GenAI |
|
| Frameworks |
|
| Databases |
|
| Dev Tools |
|
Machine Learning → Deep Learning → NLP → LLMs → LangChain → Agentic AI → MLOps → Docker
Building systems that retrieve, reason, and respond.