MANASI DUGGAL

About

Work

Amazon
|

Software Development Intern

Highlights

Developed an end-to-end debugging system for Alexa's Language Evaluation Framework (LEaF) using AWS serverless architecture, enabling automated LLM trace processing via S3 event-driven Lambda functions.

Architected infrastructure as code (IaC) with AWS CDK, ensuring secure multi-environment deployment (beta/gamma/prod) with S3, Lambda, IAM roles, and CSP-compliant security.

Built a secure file processing pipeline integrating Harmony authentication, S3, and API Gateway, automating LLM trace transformations and enabling debugging via a React-based UI.

Implemented error handling, logging, and monitoring, ensuring a secure, maintainable, and reliable system for LLM evaluation.

Accretive Technologies
|

Software Development Intern

Highlights

Built the frontend of an Employee Management application using React. Collaborated with cross-functional teams to integrate APIs, ensuring seamless data flow and real-time updates.

Improved user experience by implementing dynamic components and improving application responsiveness. Resulted in an increase in engagement and streamlined administrative processes.

Education

Indira Gandhi Delhi Technical University for Women

Bachelor of Technology

Computer Science and Engineering

Sardar Patel Vidyalaya

Class 12

Science Stream

Grade: 96.5%

Sardar Patel Vidyalaya

Class 10

Grade: 10.00

Skills

C++
C
HTML
CSS
Javascript
React
SQL
Machine Learning

Projects

AskMyAI

Summary

Designed and implemented AskMyAI, a React-based conversational chatbot leveraging the Gemini API for generating human-like responses. The app allows users to interact via natural language prompts, providing responses formatted dynamically using React Context API for state management.

Fake News Predictor

Summary

Developed a machine learning model using Python, Pandas, and scikit-learn to effectively classify news articles as real or fake, achieving 98% accuracy. Implemented text preprocessing, TF-IDF vectorization, and logistic regression, leveraging NLP techniques with NLTK for data cleaning and feature extraction.

VueFinaceTracker

Summary

Designed and developed a finance tracking application using Vue.js 3 and the Composition API. Implemented features for adding, editing, and deleting transactions with Vue Toastification for user feedback. Utilized Local-Storage for data persistence and employed custom CSS for responsive UI design.