Caroline Spoorthi David Chelli

Information Science Engineering student with skills in data analysis, machine learning, Python, SQL, and statistical methods. Experienced in academic projects involving predictive modelling, data visualization, and cloud-based analytical workflows. Seeking an opportunity in a growth-oriented company to support data-driven projects and contribute to analytical and AI-focused initiatives.

Photography, Playing Keyboard and Guitar, Digital Art, Gaming

GHCI 2025 Hackathon - Round 2 Qualifier

· Built an ethical AI dashboard for   credit‑decision explainability using an existing ML model.

· Added feature‑importance   analysis, fairness metrics, and decision simulation.

· Delivered end‑to‑end prototype aligned with XAI and fintech standards.

• 2nd Place – Spelling Bee (State Level, Karnataka)
• Merit Certification – Piano Exam, Trinity School of London
• Star Engagement Award – Malaysia University of Science and Technology

[GenAI Powered Data Analytics], [Forage-TATA]

• Developed an XAI framework to provide transparent loan decisions and risk scoring.
• Implemented counterfactual “what-if” analysis and fairness evaluation techniques.
• Integrated a pretrained ML model with a modular UI for financial decision insights.

• Created an AI system to evaluate user actions and content against defined compliance and security policies.
• Integrated rule-based and ML checks to identify non-compliant or high-risk behaviour.
• Designed alerting and reporting workflows to support governance and risk processes.
• Documented evaluation results and compliance mappings for audit readiness.

• Developed a sequence-based DL model capable of generating structured musical patterns.
• Performed audio preprocessing and feature extraction for training workflows.
• Trained neural networks to improve sequence coherence and pattern quality.
• Evaluated generated outputs and refined hyperparameters for model stability.

• Built an AI system to classify support tickets and automatically route them to appropriate departments.
• Implemented RAG-based processing to analyse ticket text, extract key details, and determine priority levels using historical data.
• Designed rule-based workflows to suggest initial resolutions and reduce manual triage load.
• Documented model behaviour, accuracy, and integration flow in alignment with ITSM standards

• Analyzed customer datasets to predict delinquency risk using AI-driven analytical techniques.
• Developed a recommendation framework to assist the Head of Collections at Geldium in selecting effective intervention strategies for at-risk customers.
• Ensured all AI-generated insights were ethical, explainable, and aligned with responsible financial decision-making practices.

• Reviewed lounge eligibility criteria and analyzed customer groupings to estimate lounge demand across different segments.
• Created a reusable lookup table and supporting justification to help British Airways apply the model to future flying schedules.
• Interpreted data trends to support operational planning and resource allocation decisions.