Shreya Sari

Shreya Sari

I ask “what if?” and then try to build it. AI enthusiast · Engineer · Professional overthinker of interesting problems

Hyderabad, India

shreyasari24@gmail.com


About

Hey! I’m an engineer who asks a lot of “what if?” questions — and usually ends up trying to build the answer. With a background in Computer Science and AI & ML, I love exploring how intelligent systems work, experimenting with new ideas, and turning random sparks of curiosity into things that actually do something useful. One day, I might be figuring out how to make a machine learn from experience. The next, I’m exploring how AI agents can work together, how to make information easier to retrieve, or how to turn an idea that sounds a little crazy into a working prototype. I’m especially drawn to the space where AI meets real-world problems. Not just building something because it uses AI, but building something that makes someone think, *“Wait, why didn't this exist already?”* I’m still learning, still experimenting, and occasionally breaking things along the way (part of the process, right?). I like understanding the *why* behind the code just as much as getting it to work. And I’m always excited by problems that make me think harder than I expected. Outside the code, I’m a curious mind with a long list of things I want to learn and an even longer list of things I want to build. I don’t have everything figured out. I just know I want to keep getting better, build things that matter, and see how far curiosity can take me. So, if you're here to see what I've been building, welcome. You're in the right place.

Skills
Vertex AI
ml
Python
Google Cloud
PyTorch
Docker
sap
JavaScript
Experience

Associate Software Engineer

EY GDS

Oct 2025 to Present
  • Contributed to enterprise-scale SAP data migration and validation at EY GDS while exploring AI-driven innovation and emerging technologies.
Projects

Multi-Agent Deep Reinforcement Learning-based Research Paper Recommendation System

A Multi-Agent Deep Reinforcement Learning (MARL) based recommendation system designed to suggest relevant research papers based on user-selected categories. The system leverages multiple agents, each trained using Deep Q-Networks (DQN), to collaboratively learn effective recommendation policies. These agents interact with the environment to optimize paper relevance and diversity, ensuring personalized and context-aware research paper suggestions.

Programming Language: Python AI/ML: Deep Reinforcement Learning
Multi-Agent Reinforcement Learning (MARL) Algorithm: Deep Q-Network (DQN) Frameworks/Libraries: PyTorch or TensorFlow
RLlib (if used) Data Processing: Pandas
NumPy Recommendation Techniques: Personalized recommendation
relevance and diversity optimization Development Environment: Jupyter Notebook
Education

Bachelors of Technology - CSE AIML

GITAM University · 8.9

Oct 2021 to Jul 2025