Sruthi Satyavarapu
Masters student with 7+ years of industry experience. Building intelligent AI systems — from modular RAG architectures to wearable health monitoring.
About Me
I'm a Masters student in Computer Science at the University of the Pacific, bringing over 7 years of industry experience in software development to the world of AI/ML.
My journey spans from bioinformatics visualization tools and enterprise B2B applications to cutting-edge Retrieval-Augmented Generation (RAG) systems and wearable health monitoring. I bridge the gap between research and production-ready AI.
Published at IEEE BigDataService 2026 (Fukuoka, Japan): A Modular Architecture for Domain Adaptive Retrieval Augmented Generation Systems. This paper presents a RAG architecture that enables swapping components through configuration alone, with session management and component-level observability built in for deployability. It shows that identical configurations behave differently across domains — making it a practical tool for researchers and practitioners to find the right pipeline for their own data.
Featured Projects
Things I've built and researched
AI Companion for Senior Citizens
AI-driven elderly care platform enabling medical data extraction, patient management, and multimodal input for improved senior care delivery.
Modular RAG Chatbot Architecture
Modular, service-oriented RAG architecture with plug-and-play swapping of embedding models, vector DBs, and LLM backends. Built end-to-end with ReactJS, FastAPI, and ChromaDB incorporating query rewriting, re-ranking, and RBAC-based access control. Integrated session-aware context management and latency instrumentation for reproducible benchmarking across domains.
Strawberry Ripeness Classifier
Supervised learning model to classify strawberry ripeness stages. Authored a white paper detailing methodology and experimental findings.
Adidas Sales Dashboard
Tableau dashboard analyzing Adidas sales data (2023–2024), uncovering regional trends and product performance insights for strategic decisions.
Weather & Music Trends Analysis
Engineered an end-to-end ETL pipeline integrating daily music charts and weather data into PostgreSQL, covering 103 cities across 6 continents over 6 months. Standardized track identities using Spotify IDs and built a Dockerized, multithreaded Essentia pipeline to replace Spotify's deprecated audio-features API. Applied PCA and KMeans clustering to group cities into 4 regional listening profiles, and analyzed how BPM, energy, danceability, valence, and musical key vary across cities and weather conditions.
Secure Hospital Management System
Role-based hospital management system for patients and doctors with OAuth2 authentication and secure API design following OWASP guidelines.
Technical Skills
Technologies and tools I work with
AI / ML & Data Science
Programming Languages
Frontend
Backend & Data Engineering
Databases
Cloud & DevOps
Security
Tools & Platforms
Work Experience
7+ years across research and industry
Research Assistant
University of the Pacific
A Modular Architecture for Domain-Adaptive Retrieval-Augmented Generation Systems
March 2025 – Present
Stockton, CA
- ◆Designed and implemented a modular RAG architecture addressing the gap between benchmarking frameworks like FlashRAG and deployment frameworks (LangChain, LlamaIndex), enabling configuration-driven experimentation without architectural changes.
- ◆Built end-to-end system using ReactJS, FastAPI, and ChromaDB incorporating query rewriting, re-ranking, session-aware interaction, and RBAC-based access control for multi-user environments. Integrated persistent storage and latency benchmarking infrastructure for reproducible cross-domain RAG evaluation.
- ◆Evaluated retrieval performance across biomedical and Sanskrit philosophical corpora using Recall@K, MRR, and RAGAS, demonstrating high domain sensitivity to component selection.
- ◆Paper published at IEEE BigDataService 2026 (Fukuoka, Japan): "A Modular Architecture for Domain-Adaptive Retrieval-Augmented Generation Systems"
Research Assistant
University of the Pacific
Position & Sleep Tracking via Wearable Sensor Data
June 2025 – Dec 2025
Stockton, CA
- ◆Developed ML models using PyTorch and Scikit-learn to analyze wearable sensor data streams for IoT-based health monitoring, improving anomaly detection accuracy by 28%.
- ◆Engineered time-series feature extraction pipelines for environmental and motion-based datasets, enhancing predictive model performance across multiple sensor modalities.
- ◆Built Flask-based RESTful services integrated with React dashboards to deliver real-time sensor insights and model outputs to end users.
AI/ML Intern
PNC, USA
Jan 2024 – July 2024
USA
- ◆Built end-to-end ML pipelines using AWS SageMaker and Airflow for financial risk systems, improving processing efficiency by 30%.
- ◆Built an ensemble fraud detection system combining XGBoost on structured transaction data with a fine-tuned LLM on unstructured transaction text into a unified fraud score, reducing detection latency by 15%.
- ◆Designed and deployed scalable ML services via Java Spring Boot microservices, Docker, and Kubernetes across cloud environments, ensuring high availability and seamless enterprise integration.
Software Developer
Tata Consultancy Services (TCS)
July 2021 – April 2023
Hyderabad, India
- ◆Developed and deployed ML models for predictive analytics on the Mars Petcare platform, improving forecasting accuracy by 25%.
- ◆Implemented secure backend systems following OWASP guidelines for enterprise applications, improving performance by 20% and reducing vulnerabilities by 35%.
- ◆Deployed and maintained scalable cloud solutions on AWS (EC2, S3, Lambda) supporting high availability.
Senior Software Engineer
Foray Software Private Limited
July 2019 – May 2021
India
- ◆Enhanced front-end inventory search functionality for Mars employees — 30% faster search, 15% increase in task efficiency.
- ◆Partnered with cross-functional teams to integrate and ensure seamless functionality.
Associate Analyst
PurpleTalk
Feb 2017 – June 2019
India
- ◆Canya App: Developed features enabling peer-to-peer digital services on Android.
- ◆Cuckuu App: Built and unit tested a social networking alarm app. Received Best Performance Award.
- ◆SILA: Developed and maintained the payment & invoice module within SILA Attendance Management. Received Best Employee Award.
Research Associate / Intern
Excellera (GVK Bio)
June 2015 – Dec 2016
India
- ◆Developed an automated visualization tool to map disease associations and biological pathways for target gene IDs, improving workflow efficiency by 25%.
Education
Academic background and qualifications
Master of Science in Computer Science
University of the Pacific
Stockton, CA · Aug 2024 – June 2026
Master of Science in Medical Software Technology
Manipal Institute of Technology
Manipal, India · Aug 2014 – May 2016
Bachelor of Engineering in Bio Medical Engineering
Jawaharlal Nehru Technological University
India · Aug 2010 – May 2014
Get In Touch
Open to opportunities, collaborations, and research discussions
Let's Connect
I'm actively seeking AI/ML engineering roles and research collaborations. Whether you have an opportunity, a project idea, or just want to discuss the future of AI — I'd love to hear from you.