Ramya V.

Key Skills

Software

Other

AWS SageMaker

Top Subject Matter

Employee feedback analytics for organizational insights

Federated learning (fairness) on edge devices

Top Data Types

Text

Image

Top Task Types

Text Summarization

Classification

Fine-tuning

Freelancer Overview

AI Intern, Ford Motor Company. Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other and AWS SageMaker. Education includes Bachelor of Technology, Shiv Nadar University (2026) and Higher Secondary Certificate (Grade 12), Mahatma Montessori School (2022). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Text Summarization, Classification, and Sentiment Analysis.

Labeling Experience

AI Intern, Ford Motor Company

OtherTextText SummarizationClassification

Built an automated Voice of Employee analytics engine using LangChain and LLMs to transform large-scale global employee feedback into actionable insights. Implemented a modular pipeline for multi-level theme classification, sentiment analysis, and insight summarization over textual responses. Focused on converting unstructured textual feedback into structured, decision-ready outputs for organizational stakeholders.• Ingested and processed large volumes of employee feedback text for analysis.• Produced theme and sentiment predictions for multi-level categorization.• Generated summarized insights from analyzed text using LLMs.• Reduced manual review effort by automating interpretation workflows.

2026 - 2026

AI Research Intern, IIT Madras

AWS SageMakerFine-tuning

Worked on EmbracingFL, a fairness-driven federated learning approach for heterogeneous edge devices involving ML training across parallel clients. Focused on scaling distributed training using GPU acceleration to achieve strong benchmark accuracy on CIFAR-10 with ResNet-20. Contributed to training and evaluation processes aimed at improving fairness and performance under non-uniform data conditions.• Orchestrated federated learning across heterogeneous edge environments.• Enabled GPU-accelerated parallel client training for scalability.• Trained/evaluated a ResNet-20 model on CIFAR-10 for CIFAR-10 accuracy reporting.• Supported fairness-driven modeling for non-IID data settings.

2025 - 2025

Education

Work History