Sep 2025 – Dec 2025
QwenNET
- Used the Qwen LLM to automate detection of complex system anomalies, with model feedback loops and agentic understanding of raw data streams.
- Designed an end-to-end data pipeline that ingests raw network traces and turns them into structured corpora for anomaly detection at scale.
PythonPyTorchQwen LLMScapy
Feb 2025 – May 2025NYU team
Multilingual Music Recommendation Model
- Architected data pipelines over the Million Playlist Dataset, using MinIO (S3-compatible) and Docker to simulate cloud-native environments with reproducible workflows.
- Built pipelines that process lyrics and generate positive training pairs for personalized, cross-language music discovery.
PythonPyTorchDockerMinIO
Aug 2023 – Apr 2024Capstone
AI-Powered Task Automation Agent
2nd place · Annual Project Competition, Dept. of Computer Engineering, NMIMS
- Designed an AI-first consumer interface using Hugging Face models and React to automate complex user workflows.
- Built the autonomous agent for scalability and flexibility, adapting to diverse user needs in real time.
PythonReactTypeScriptVector DB