VR

Venkata sai rohith

B.E. student in Artificial Intelligence and Machine Learning · Aspiring AI/ML Engineer

Bengaluru, Karnataka, India

@venkata_sai_rohith

0 followers

B.E. in Artificial Intelligence and Machine Learning · Graduating 2027

About

Artificial Intelligence and Machine Learning undergraduate at Sir M. Visvesvaraya Institute of Technology, Bengaluru, with hands-on project experience in RAG systems, NLP, and resume analysis applications. Comfortable building backend-powered AI workflows using Python, FastAPI, LangChain, and related machine learning tools.

Education

Skills

Tools / apps / platforms

<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.
Git Groq GitHub Ollama PostgreSQL Hugging Face

Projects

  • Resume analysis app
    Python, Streamlit, spaCy, Sentence Transformers, pdfplumber

    Developed a resume analysis app that compares resume content with job descriptions using NLP and semantic similarity. Implemented PDF text extraction, text preprocessing, keyword extraction, and transformer-based embeddings for matching. Built a skill-gap analysis module to identify technical skills from job requirements that are missing from a resume. Created a Streamlit interface for resume uploads, match-score visualization, and feedback. Built stateful agent workflows with LangGraph and experimented with tool-based execution. Developed an MCP server to expose tools to LLM-based agents and explored local LLM integration using Ollama.

  • Multi-paper RAG application
    Python, FastAPI, LangChain, FAISS, BM25, Hugging Face, Groq, PyMuPDF

    Built a multi-paper RAG application for querying and summarizing research papers using an LLM-powered question-answering pipeline. Developed PDF ingestion, text extraction, semantic chunking, embedding generation, and vector indexing through a FastAPI backend. Implemented dense semantic retrieval and BM25 keyword retrieval, combining results using Reciprocal Rank Fusion (RRF). Built an evaluation suite for planner routing, retrieval, answer faithfulness, and summarization; achieved 100% routing accuracy (25/25), keyword Hit@5 of 1.00, MRR of 0.955, and QA faithfulness of 4.67/5. Evaluated summaries for coverage, coherence, and conciseness, scoring 5/5, 5/5, and 4/5 respectively using an LLM-as-judge. Added source citations to support traceable answers and designed independent vector stores for multiple research papers.

Courses & certifications

  • AI Foundations Associate · Oracle University · 2025
  • Google AI Essentials · Google · 2025

Explore the Broxer community

Online · instant AI help
Broxer