PS

Pavithra S

M.Tech Software Engineering student · AI & Machine Learning practitioner

Vellore, Tamil Nadu, India

@pavithra_s1

0 followers

Integrated M.Tech in Software Engineering at Vellore Institute of Technology (VIT) · Graduating 2027

About

Integrated M.Tech Software Engineering student with hands-on experience in machine learning, deep learning, NLP, and RAG systems. Builds Python-based AI workflows, REST APIs, and cloud-backed services using modern ML frameworks and backend tooling.

Experience

  • AI Intern
    Infosys Springboard
    Nov 2025 – Present

    Built an XGBoost classification solution for geospatial analytics using data processing and feature engineering, achieving a 97.71% weighted F1-score. Automated Python ETL pipelines using the OpenWeatherMap API, reducing processing latency by 30% while maintaining data integrity. Evaluated model performance, troubleshot pipelines, used Git version control, and documented technical findings.

Education

  • Integrated M.Tech in Software Engineering
    Software Engineering · 2027
  • Class XII
    Sri Ram Dayal Khemka Vivekananda Vidyalaya Junior College
  • Class X
    Sri Ram Dayal Khemka Vivekananda Vidyalaya Junior College

Skills

Tools / apps / platforms

1 to 2 years 1 to 2 years A year or two of steady use.
Git
<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.
Neo4j Chroma Docker Amazon Web Services AWS

Projects

  • Multilingual Emotion Detection (Regional Languages)
    XLM-RoBERTa, PyTorch, TensorFlow

    Tamil/Telugu text classification system achieving 98% accuracy on 5,000+ records with an end-to-end deep learning pipeline.

  • Foundation Shade Matcher
    OpenCV, HSV, Node.js, Express

    Image-based foundation recommendation system that extracts skin-colour information from an uploaded photo and matches it against a foundation dataset.

  • Enterprise RAG with Knowledge Graph and Hybrid Vector Retrieval
    Neo4j, FAISS, Chroma, LangChain, LangGraph, FastAPI, Qwen2.5

    End-to-end RAG system using knowledge graphs, vector search, semantic embeddings, hybrid retrieval, and reranking. Built backend services for document ingestion and query workflows and achieved a 99.4% automated test pass rate.

Courses & certifications

  • Artificial Intelligence Primer · Infosys Springboard
  • Quantum Computing Fundamentals · Qiskit
  • Machine Learning with Python · IBM

Explore the Broxer community

Online · instant AI help
Broxer