Lahari image

Lahari Balantrapu

also known as IvyAeri

C:/portfolio/about/Lahari:

Lahari.age()

Lahari.location()

"London, UK"

Lahari.education()

"MSc Artificial Intelligence and Machine Learning student at University of Birmingham."

Lahari.skills()

["PyTorch", "TensorFlow", "Scikit-learn", "OpenCV", "Keras", "NLTK", "SpaCy", "BERT", "GPT", "Pandas", "NumPy","Weights & Biases", "AWS", "Google Cloud", "Prompt Engineering", "OpenAI", "Claude","Geocoders"]

Lahari.experience()

Scrub_AI: {
    "period": "2020 - present",
    "projects": [{
        {"Enhance AI models: {
            "Key Achievements": "Focused on optimising and enhancing the performance of existing AI models.",
            "Technologies Used": ["Python", "Scikit-learn", "Pytorch","Weights & Biases","s3"]
        }
        },
        "Exploring Large Language Models (LLMs)": {
            "Key Achievements": "Investigated the capabilities and applications of LLMs, improving the performance.",
            "Technologies Used": ["GPT-4", "Claude", "Prompt Engineering"]
        }
        },
        {"Setting Up Pipelines for LLMs": {
            "Key Achievements": "Developed scalable pipelines, integrated LLMs into inference.",
            "Technologies Used": ["Open-source LLM models","Python"]
        }
        },
        {"Streamlining Data Labelling Pipelines": {
            "Key Achievements": "Enhanced efficiency and accuracy of data labelling processes, implemented Label Studio.",
            "Technologies Used": ["JavaScript", "React", "Label Studio","Python"]
        }
        },
}

University_of_Birmingham: {
    "period": "2018 - 2020",
    "projects": [{
        "Medical Imaging and Diagnostics": {
            "Key Achievements": "Applied AI techniques to medical imaging, improved diagnostic accuracy by 15%.",
            "Technologies Used": ["OpenCV", "Keras", "TensorFlow","PyTorch", "Scikit-learn"]
        }
        },
}

Lahari.Display(Projects)

// successfully compiled, scroll down to view.

My Portfolio

This portfolio is a device-friendly website of my resume that can be viewed on all the screen sizes.

HTML Bootstrap AOS CSS

Prediction of Skin Cancer

A machine learning model that predicts the stage of skin cancer based on the image of the affected area.

Python OpenCV Scikit-learn Machine Learning Tkinter

BackPropagation and SoftMax on Fashion MINST

This project is about classifying the fashion MINST data set by using the concepts of BackPropagation and SoftMax. The coding work was done based on the mathematical concepts and without using any inbuilt functions. Testing has been done to find the optimal hyperparameters.

Python Pandas NumPy Machine Learning Matplotlib

Semantic Segmentation of MR images

This project is aimed to create a network to segment MR images of hearts into 4 sections. A U-Net based architecture was chosen for the task. Experiments were performed to optimise the hyperparameters of the network and compare different architectures. The modified U-Net architecture was found to perform better than other architectures tested.

Python CNN Jupyter Machine Learning PyTorch OpenCV Matplotlib

University Admission Prediction

Linear regression along with different optimisation techniques have been implemented without using the inbuilt functions to predict the admission of a student based on several categories such as IELTS, CGPA, SOP, etc.

Python Pandas NumPy Machine Learning Matplotlib

"Education, I believe, is a never-ending process and so is my portfolio, it keeps getting better."

I am a self-motivated individual always thriving to learn new things, be it academic or extra-curricular. Feel free to drop a message to any of my socials below to further discuss any possible collaborations and employment opportunities.