Hyderabad, Telangana, India
3K followers 500+ connections

Join to view profile

About

Hello,
I am working as a Solutions Architect at AWS. I am a published author for the…

Articles by Pavan Kumar Rao

Activity

Join now to see all activity

Experience & Education

  • Amazon Web Services (AWS)

View Pavan Kumar Rao’s full experience

See their title, tenure and more.

or

By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.

Licenses & Certifications

Join now to see all certifications

Publications

  • Fine-tune Meta Llama 3.2 text generation models for generative AI inference using Amazon SageMaker JumpStart | AWS Machine Learning Blog

    Amazon Web Services

    See publication

  • Fine-tune Llama 3 for text generation on Amazon SageMaker JumpStart

    AWS

    See publication

  • Inference AudioCraft MusicGen models using Amazon SageMaker

    Amazon Web Services

    See publication

  • Inference Llama 2 models with real-time response streaming using Amazon SageMaker

    Amazon

    See publication

  • Getting Started with V Programming

    Packt Publishing

    A new language on the block, V comes with a promising set of features such as fast compilation and interoperability with other programming languages. This is the first book on the V programming language, packed with concise information and a walkthrough of all the features you need to know to get started with the language.

    The book begins by covering the fundamentals to help you learn about the basic features of V and the suite of built-in libraries available within the V ecosystem…

    A new language on the block, V comes with a promising set of features such as fast compilation and interoperability with other programming languages. This is the first book on the V programming language, packed with concise information and a walkthrough of all the features you need to know to get started with the language.

    The book begins by covering the fundamentals to help you learn about the basic features of V and the suite of built-in libraries available within the V ecosystem. You'll become familiar with primitive data types, declaring variables, arrays, and maps. In addition to basic programming, you'll develop a solid understanding of the building blocks of programming, including functions, structs, and modules in the V programming language.

    As you advance through the chapters, you'll learn how to implement concurrency in V Programming, and finally learn how to write test cases for functions. This book takes you through an end-to-end project that will guide you to build fast and maintainable RESTful microservices by leveraging the power of V and its built-in libraries.

    By the end of this V programming book, you'll be well-versed with the V programming language and be able to start writing your own programs and applications.

    See publication

  • Engineering Solutions To Combat Climate Change

    Institute Of Engineers India

    A research writing on how the global warming can be reduced by Engineering means.

Projects

  • Keyspaces CDC Streams to S3

    -

    Stream Amazon Keyspaces Change Data Capture (CDC) events to Amazon S3 using the AWS Keyspaces Streams Kinesis Adapter.

  • Inference AudioCraft MusicGen models using Amazon SageMaker

    -

    With the ability to generate audio, music, or video, generative AI models can be computationally intensive and time-consuming. Generative AI models with audio, music, and video output can use asynchronous inference that queues incoming requests and process them asynchronously. Our solution involves deploying the AudioCraft MusicGen model on SageMaker using SageMaker endpoints for asynchronous inference. This entails deploying AudioCraft MusicGen models sourced from the Hugging Face Model Hub…

    With the ability to generate audio, music, or video, generative AI models can be computationally intensive and time-consuming. Generative AI models with audio, music, and video output can use asynchronous inference that queues incoming requests and process them asynchronously. Our solution involves deploying the AudioCraft MusicGen model on SageMaker using SageMaker endpoints for asynchronous inference. This entails deploying AudioCraft MusicGen models sourced from the Hugging Face Model Hub onto a SageMaker infrastructure.

  • Amazon SageMaker Llama 2 Inference via Response Streaming

    -

    Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. The fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency, and knowledge tests.

    This repo helps customers looking to have faster response times in the form of TTFB and thus reduce the overall perceived latency…

    Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. The fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency, and knowledge tests.

    This repo helps customers looking to have faster response times in the form of TTFB and thus reduce the overall perceived latency. The streaming support is possible with the latest announcement Sagemaker Real-time Inference now supports response streaming.

    The samples covers notebook recipes on how to implement Response Streaming SageMaker Endpoints for Llama 2 LLMs. These models were deployed using the Amazon SageMaker Deep Learning Containers HF TGI and DLC for LMI. To be precise, these are DLC for Large Model Inference and the recently announced Hugging Face DLC powered by Text Generation Inference.

    This repo covers Deploy and Inference Llama 2 Models on SageMaker via Response Streaming.

    https://github.com/aws-samples/amazon-sagemaker-llama2-response-streaming-recipes

  • FastAPI Deployment Tutorials

    -

    Tutorials that demonstrate how to deploy FastAPI on various cloud platforms as well as on-prem.

  • Cyclops - A Computer Vision based Text Recognition & Object Detection App.

    -

    Cyclops, which targets elimination of manual efforts in validating the scanned documents using Cognitive Computing.

    The main aim of this project is to recognize the Hand Written, Optical Characters and custom objects like images of stamps from the scanned documents and validate it against the ground truth information stored in the database.

    The challenging part of this project was implementing a validation logic that identifies the presence of data present in the ground truth…

    Cyclops, which targets elimination of manual efforts in validating the scanned documents using Cognitive Computing.

    The main aim of this project is to recognize the Hand Written, Optical Characters and custom objects like images of stamps from the scanned documents and validate it against the ground truth information stored in the database.

    The challenging part of this project was implementing a validation logic that identifies the presence of data present in the ground truth database against the scanned document using the naive approach of Regex based text classification technique to identify frequent patterns that sufficed our requirement for this application. In addition to this, the app can also recognize the type of document being processed based on pre-configured rules by using Document Similarity measure 'Cosine similarity'.

    This application is capable of
    OCR
    Hand Written Text Recognition (HTR)
    Spell Check and Auto Correction
    Object Recognition
    Document Similarity

    We have used this application for Computerized vehicle registration to automate the auditing process of Vehicle Registration documents which come in the form of scanned images. It is noticed that on an average each document takes 15-45 secs. 15-20 secs for OCR/HWT recognition + validation and 30-40 secs if OCR+ HTR+ Object Recognition + Document Similarity + validation applied simultaneously.

    Tech Stack: Python, C#, .NET, Azure
    Domain: Natural Language Processing, Cognitive Computing, Custom Vision

  • Facial Feature Recognition and Masking using Haar feature-based cascade classifiers

    -

    There are many situations where medical surgeons and students in the field of medical research happen to work with facial image datasets of the patients and it is quite necessary to protect the identity of the patient. When dealing with large datasets it is often a cumbersome task to manually blur facial parts using traditional methods like Photoshop or MS Paint. This Project aims at intelligently blurring the facial features like eyes, nose, mouth given an image or set of images without manual…

    There are many situations where medical surgeons and students in the field of medical research happen to work with facial image datasets of the patients and it is quite necessary to protect the identity of the patient. When dealing with large datasets it is often a cumbersome task to manually blur facial parts using traditional methods like Photoshop or MS Paint. This Project aims at intelligently blurring the facial features like eyes, nose, mouth given an image or set of images without manual effort. The approach uses computer vision library in python named OpenCV and for face and facial parts detection is done using Haar Feature-based Cascade Classifiers.

    See project

  • kNN from the Scratch & Comparison with Scikit Learn's kNN Implementation

    -

    This project aims to implement kNN algorithm from the scratch using Python programming language and compare its performance with the scikit learn's implementation of kNN in terms of accuracy and run times. The custom implementation of kNN had an average accuracy of 47.1% in contrast to scikit learn's kNN which has an average accuracy of 53% approximately , and custom kNN had an average run-time of 1.3 ms in contrast to scikit learns knn which has an average run-time of 0.003 ms approximately…

    This project aims to implement kNN algorithm from the scratch using Python programming language and compare its performance with the scikit learn's implementation of kNN in terms of accuracy and run times. The custom implementation of kNN had an average accuracy of 47.1% in contrast to scikit learn's kNN which has an average accuracy of 53% approximately , and custom kNN had an average run-time of 1.3 ms in contrast to scikit learns knn which has an average run-time of 0.003 ms approximately on a non-realistic dataset created on the fly using pandas and numpy.

    See project

  • Opinion Mining based on Product Reviews

    -

    Opinion Mining using Python, Natural Language Processing using NLTK
    This project is about finding frequent aspects and opinions from the product reviews of digital camera. The opinion Mining is performed by basic text pre-processing using Opinion Finder tool, together with Natural Language processing and Python.

    See project

  • TV Show Recommender System

    -

    This project aims to recommend an user the most relevant shows based on user-user and show-show collaborative filtering based on his past show preference history.

    See project

Honors & Awards

  • Most Valuable Player - Q2 2025 @ AWS

    Amazon Web Services

    Most Valuable Player - Q2 2025 @ AWS

  • SMGS India #OneAWS award

    AWS

    For outstanding work as a part of Gen AI Focus Group

  • Sri Bharata Ratna Mokshagundam Visweswaraya Award

    Institue of Engineers India, Hyderabad

    A research writing on how the global warming can be reduced on Earth by Engineering means.

  • College Topper in Probability Theory and Stochastic Process

    -

    Secured highest score 91% in Probability Theory and Stochastic Process and stood all time top in the college.

  • "A" Team Award - May 2020, Aug 2021

    -

  • "A" Team Award - Nov 2020

    -

Languages

  • English

    Full professional proficiency

  • Marathi

    Native or bilingual proficiency

  • Hindi

    Full professional proficiency

  • Telugu

    Native or bilingual proficiency

Recommendations received

6 people have recommended Pavan Kumar Rao

Join now to view

More activity by Pavan Kumar Rao

View Pavan Kumar Rao’s full profile

  • See who you know in common
  • Get introduced
  • Contact Pavan Kumar Rao directly
Join to view full profile

Other similar profiles

Explore more posts

Explore top content on LinkedIn

Find curated posts and insights for relevant topics all in one place.

View top content

Add new skills with these courses

See all courses