profile-pic

Shaurya Raina

Generative AI Engineer with a passion for leveraging cutting-edge AI technologies to solve real-world problems. Skilled in building and deploying large language models (LLMs) for tasks like content generation, automation, and AI-driven solutions. With expertise in machine learning, data analysis, and prompt engineering, I’m dedicated to driving innovation and pushing the boundaries of AI to create intelligent, scalable systems.

  • Role

    Gen AI Engineer

  • Years of Experience

    3 years

Skillsets

  • Data Cleaning
  • Artifact detection
  • rag
  • Prompt Engineering
  • Power BI
  • OpenAI API
  • Notion
  • multimodal AI
  • Model evaluation
  • LLMs
  • HuggingFace
  • Funnel Optimization
  • fine-tuning
  • Event Tracking
  • Python
  • TensorFlow
  • Superannotate
  • Streamlit
  • PyTorch
  • LangChain
  • Labelbox
  • Jira
  • Gradio
  • Google Analytics
  • FastAPI
  • Cvat
  • SQL

Professional Summary

3Years
  • Aug, 2024 - Present1 yr 7 months

    Gen AI Engineer

    Nvidia
  • Aug, 2023 - Aug, 20241 yr

    Data Analyst

    Careers 360

Work History

3Years

Gen AI Engineer

Nvidia
Aug, 2024 - Present1 yr 7 months
    Fine-tuning large language models (LLMs) and optimizing AI pipelines using Python, TensorFlow, and PyTorch. Contributing to Drive-AV (Autonomous Vehicles) project with Mercedes-Benz, focusing on Level 14 autonomy and safer self-driving systems. Part of NeMo Generative AI Data Factory (Dense Caption Team) enhancing multimodal data and intent recognition. Evaluated AI-generated videos for visual, spatial, and physical artifacts, improving generative model quality and realism. Collaborated with the AI Research Team to refine multimodal artifact detection and feedback systems.

Data Analyst

Careers 360
Aug, 2023 - Aug, 20241 yr
    Managed user data from various educational products and CTAs on the website using SQL, employing Google Event Tracking for detailed analysis. Cleaned and analyzed data to identify user patterns, facilitating data-driven decision-making. Proficient in traffic analysis and user behavior identification using Google Analytics, contributing to funnel creation and loophole identification based on user-driven data. Enabled lead generation for clients by creating and optimizing funnels based on product-derived data. Identified valuable data for specific clients, contributing to lead generation as a major business vertical for the company. Managed tasks using Jira, actively participated in scrums, and conducted routine meetings with internal team and tech teams. Successfully managed the migration of three projects (Review Pages, College Listing Pages, and Search Functionality of the website) and maintained ongoing product enhancements. Acted as the Single Point of Contact (SPOC) for five critical products, overseeing regular trends identification and resolution of issues. Applied analytical insights to contribute to the development of efficient product features, enhancing overall product functionality and business results. Presented data-based insights through visualizations using Google Looker Studio, Power BI for effective stakeholder communication. Improved identification of system loopholes and bottlenecks, contributing to better decision-making.

Major Projects

4Projects

Air Quality Index Prediction (Machine Learning Course OE6b Project)

    Led a team in developing a hybrid machine learning model to predict Air Quality Index by analyzing chemical pollutant levels using regression and classification models. Achieved a final score of 58.16%.

Multimodal Prompt-to-Video Generation Evaluator

    Evaluated AI-generated videos from text prompts, critiqued realism, motion physics, and object consistency. Developed a framework for artifact taxonomy and feedback loop for model improvement using Sora/Runway ML, Python, and CVAT.

Personalized Chatbot Using Retrieval-Augmented Generation (RAG)

    Built a domain-specific chatbot using custom documents and real-time search with RAG. Implemented memory, summarization, and embedding techniques for context-aware responses using LangChain, OpenAI API, FAISS, and Streamlit.

AI-Powered News Headline Generator with Emotion Tuning

    Designed a system that generates emotionally tuned news headlines for the same article using LLMs and fine-tuned sentiment classifiers with GPT-4 and Hugging Face Transformers.

Education

  • BE Graduate(Honours, Data Science)

    Bharati Vidyapeeth Women's College of Engg (2024)

Certifications

  • JP Morgan Chase Software Engineering