Skip to content

AIWithShrey/AIWithShrey.github.io

Repository files navigation

All about me

Education

  • B.S, Computer Science | Missouri University of Science and Technology (Fall 21' - Fall 25')
  • Diploma, International Baccalaureate | Indus International School, Hyderabad (May 2021)

Work Experience

NVIDIA University Ambassador (Dec 2022 - Present)

  • Missouri University of Science and Technology's youngest NVIDIA University Ambassador (Usually given to professors)

NVIDIA University Ambassador. (Dec 2022 - Present)

  • Missouri University of Science and Technology's youngest Deep Learning Institute Certified Instructor and Univesity Ambassador.

  • Delivers a semesterly workshop on Fundamentals of Deep Learning. Free for students and certification upon passing the assessment.

Product Engeering Intern @ InfraCloud Technologies (June 2024 - August 2024)

  • Developed a platform for AI developers to deploy applications on a GPU-enabled cluster, enabling a cost-efficient, pay-as-you-go approach using high-performance infrastructure.

  • Automated platform-building processes using:

    • Network booting: Simplified setup for multiple machines.
    • Orchestration: Deployed Kubernetes to manage cluster operations.
    • AI-stack integration: Implemented Hugging Face's Text Generation and Embedding Inference stack, and explored NVIDIA NIMs for faster, more efficient workflows.
  • Generative AI PoC: Built a proof-of-concept involving a chatbot driven by a Generative AI agent equipped with tools to turn conversations into actions, such as searching the web or Stack Overflow and analyzing the team's Slack messages for real-time insights. Powered by Retrieval-Augmented Generation (RAG) and Function-Calling, InSightful (https://github.com/infracloudio/insightful) delivers actionable results with minimal input.

  • Collaborated with an incredible team of talented professionals whose mentorship was invaluable in refining my technical and problem-solving skills.

Summer Intern @ LegalCraft Solutions (June 2023 - August 2023)

  • Worked closely with the development team to build a chatbot that lawyers can use to interact with their private documents pertaining to a specific case. LegalCraft's product, Lexel, is a platform for lawyers to collaborate and work on various documents together.

  • The chatbot helps in acquiring important details regarding the case with a query in the simple natural language form. The chatbot excels in summarising, producing a timeline of events, extracting important entities and their corresponding details as well.

  • Utilized Microsoft Azure (Azure OpenAI and Azure Cognitive Search Services) to cut operational costs and increase the overall security concerning the customers and their data.

Certifications

  • TensorFlow Certified Developer
  • Linux Foundation Certified Systems Administrator
  • Certified Kubernetes Application Developer
  • NVIDIA DLI Certified Instructor

Projects

BLIP2 Fine-Tuning on NYT unmasked dataset

  • Used Bootstrapping Language-Image Pre-training Version 2 (BLIP2) model to predict the news category on a given NYT image.

  • Fine-Tuned the model to generate an NYT-style headline for a given image, making it a "headline generator".

  • Image captioning with the flavours of NYT headlines.

    GitHub Repo

Celebrity Song Player

  • Building an application using TensorFlow for image recognition and Spotify API to identify Bollywood actors/actresses and play a random song from one of their movies.

  • Listen to the songs of your favorite actors and actresses by simply pointing your smartphone camera at their image!

  • Currently in-progress, curating the dataset is the biggest challenge to this project.

Pneumonia detector

  • Used a pre-trained image recognition model (VGG16) on TensorFlow Hub and fine-tuned it to find traces and the presence of Pneumonia in a given x-ray scan of the lungs.

  • The model returns a boolean (True or False) value to show the presence of Pneumonia, making this a binary classification project.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published