Skip to content
View Yash3561's full-sized avatar
  • Student
  • Newark, New Jersey
  • 09:33 (UTC -05:00)
  • LinkedIn in/ygc2003

Highlights

  • Pro

Block or report Yash3561

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Yash3561/README.md

Yash Chaudhary | Generative AI Engineer

MS in AI candidate at the New Jersey Institute of Technology with a focus on building and benchmarking production-grade Generative AI systems. My work spans the full development lifecycle, from implementing cutting-edge research to engineering scalable, high-performance AI applications.


🛠️ Core Competencies & Tech Stack

PyTorch Transformers vLLM Docker Python AWS

  • GenAI & LLMs: RAG Architecture, LLM Fine-Tuning (PEFT/LoRA), Autonomous Agents (ReAct), Vector Databases (FAISS, ChromaDB), LangChain
  • ML & Data Science: Scikit-Learn, Pandas, NumPy, XAI (SHAP, LIME), Computer Vision
  • Engineering & DevOps: Git, REST APIs (Flask/FastAPI), Streamlit, SQL

🚀 Flagship Projects

Project Description Key Technologies
Project Chimera A resilient, multimodal autonomous AI agent with sandboxed tool-use (file system, code interpreter, web) and a human-in-the-loop fallback. Python, vLLM, Transformers, ChromaDB, Florence-2, Whisper
MedQuery-RAG An end-to-end medical RAG system powered by vLLM for 5-10x inference throughput, featuring an intent classifier and cross-encoder re-ranker. Python, vLLM, RAG, FAISS, Streamlit, Cross-Encoders
Project CLAIRE A benchmark study on continual learning, empirically demonstrating the trade-off between factual retention and linguistic coherence in LLMs. Python, PyTorch, PEFT/LoRA, Llama-3.1, LLM-as-a-Judge
Self-Align for Go An implementation of the Self-Align data generation pipeline to create a high-quality, validated instruction-response dataset for the Go language. Python, StarCoder2, vLLM, Go, Data Pipelines

Pinned Loading

  1. MedQuery-RAG MedQuery-RAG Public

    Python 1

  2. Project_Chimera Project_Chimera Public

    Python 1

  3. Project_CLAIRE Project_CLAIRE Public

    Jupyter Notebook 1

  4. SelfAlign_Project SelfAlign_Project Public

    Jupyter Notebook 1 1