Machine Learning Software Developer

SHRUTIKA SUNIL PARAB

Jersey City, NJ ***** 551-***-**** ****************@*****.*** LinkedIn GitHub Portfolio

Summary

AI/ML Engineer with 5+ years of experience building scalable machine learning solutions and intelligent platforms across healthcare, recruitment, and fintech sectors. Skilled in Python, PyTorch, TensorFlow, Hugging Face, and LangChain. Proven ability to design ML pipelines, deploy deep learning models, and automate workflows. Experience includes integrating model inference systems, optimizing performance, and contributing to reproducible, production-ready ML infrastructure. Passionate about driving impact through AI innovation in fintech environments.

Experience

AI Intern: House of Good Deeds Organization, New York, NY Jan 2025 - Present

Build and deploy an NLP-powered chatbot using LangChain to automate volunteer and donor engagement

Develop onboarding guides and refine workflows based on user feedback to enhance cross-functional efficiency

Software Developer AIML Engineer: Reliance Jio Platforms Limited, Mumbai, India June 2020 – Jan 2023

Increased recruitment efficiency by 35% by designing a sentiment and emotion recognition-based video interview assistant using Python and machine learning models

Designed and deployed deep learning and computer vision models for real-time image and video analysis, utilizing TensorFlow and OpenCV, reducing processing latency by 20%

Developed an AI tool for extracting high-quality Q&A pairs from textual context using transformer-based models, Integrated Hugging Face models to generate context-aware questions and answers

Queried and analyzed large-scale datasets with SQL database-Python pipelines, providing insights that optimized user engagement strategies by 15% for the Data Science career fair

Software Development Engineer: Reliance Jio Platforms Limited, India. Sep 2018 – June 2020

Created an AI-driven recruitment bot combining OCR, sentiment analysis, and facial recognition, improving candidate screening efficiency by 25% for the Enterprise platform

Launched "HerCircle", a global women empowerment platform, using customer segmentation and clustering algorithms for classification to personalize experiences for over 100,000+ users

Implemented recommendation engines and NLP-based solutions for chatbots, increasing user engagement and satisfaction by 20%

Education

Yeshiva University Katz School of Science and Health, New York, NY: Master's in Artificial Intelligence May 2025

Mumbai University, Mumbai, India: Bachelor of Technology in Computer Engineering June 2018

Technical Skills

Languages & Frameworks: Python, SQL, NumPy, Pandas, C#, .NET Core, R

ML Frameworks: PyTorch, TensorFlow, Keras, Hugging Face, FastAPI, Flask, AWS, Databricks, Azure, Scikit-Learn

Tools & Infra: Git, Docker, PowerBI, Postman, LangChain, Streamlit, Jupyter, Workflow Automation Tools

Domains: ML Ops, NLP, Deep Learning, Gen AI, Transformers, Model Deployment, AI Infrastructure, RAGs

Projects

Dog Cardiomegaly Precision Sep 2024

Built an EfficientNet-based model to estimate vertebral heart score (VHS), achieving 95% precision

Applied transfer learning and hyperparameter tuning to improve canine diagnostic accuracy

Convolutional Neural Network for Dog Heart Disease Assessment Nov 2024

Designed a custom CNN architecture using TensorFlow to enhance heart disease detection

Employed advanced dataset augmentation to boost model robustness, improving accuracy by 20%

QnA Extractor Using Transformer Models March 2025

Built an NLP pipeline using Hugging Face transformers to generate Q&A pairs from documents, facilitating intelligent document parsing. Integrated into open-source chatbot systems

Emphasized experimental design and active learning to improve retrieval accuracy

DeepFakeSheild May 2025

Built a real vs. fake image classifier with Streamlit and CNNs, achieving 87% test accuracy on facial datasets, including stylized AI-generated images, which helps fraud detection in posts

Enabled explainability using Hugging Face LLMs, improving user trust and enhancing interpretability

Contact this candidate