Humaira

About

Applied data scientist and data analyst with hands-on experience building NLP, LLM, and analytics solutions across the full AI lifecycle. Strong focus on scalable, maintainable systems using Python, machine learning, and modern data workflows.

Experience

• Program Associate - Data Analyst
  CEGIS Foundation · Chennai
  Jan 2026 – Present

  Analyse complex data, develop models, and generate actionable business insights.

Education

• 2025 CT
  Bachelor of Computer Applications (BCA)

  CTTE College for Women

  Aug 2022 – Apr 2025

  Aug 2022 – Apr 2025
• 2025 II
  Advanced Artificial Intelligence and Machine Learning

  IIT-M GUVI

  Awarded by IIT-M

  Aug 2024 – Feb 2025

  Aug 2024 – Feb 2025

Skills

Tools / apps / platforms

Projects

• Data-Driven Stock Analysis Nifty50
  Python, Data Wrangling, Feature Engineering, Pandas, SQL, Power BI, Streamlit

  Built an end-to-end analytics platform to analyze Nifty-50 stock performance using data wrangling, statistical analysis, and feature engineering techniques. Identified top-performing, volatile, and sector-wise trends and presented insights through interactive dashboards. Delivered a production-ready visualization interface enabling stakeholders to explore market trends.

• PowerPulse: Household Energy Consumption Forecasting
  Python, Time-Series Analysis, ARIMA, Feature Engineering, Data Preprocessing, Model Evaluation

  Developed a time-series forecasting system to predict household energy consumption using ARIMA and machine learning models on smart-meter data. Performed data cleaning, transformation, and feature engineering to improve model performance, achieving an R² score of 0.89. Generated insight-driven visualizations to support data-driven decision-making in energy usage optimization.

• RAG-Powered PDF Chatbot with Citation Support
  Python, NLP, LLMs, RAG, Hugging Face, LangChain, Vector Databases, FastAPI, Docker

  Designed and built a RAG-based conversational system enabling accurate question-answering over large document collections using embeddings and vector search. Implemented an end-to-end pipeline including document ingestion, text chunking, embedding generation, semantic retrieval, and LLM-based response synthesis. Exposed inference through REST APIs and deployed a lightweight UI for real-time interaction, with focus on response accuracy, latency, and scalability.