Amit Kumar

Experience

• 

  INeuron.ai

  Aug 2020 - Dec 2020

  Intern

  • Enhanced driver comfort and reduced search time by assisting a team of 5 in developing a parking spot detection system.• Increased model efficiency by 43.7% by implementing a computer vision solution for unmarked parking spots.• Developed skills in teamwork, meeting deadlines, effective collaboration, and conflict resolution

• 

  National Institute of Technology , Patna

  Jan 2021 - Dec 2021

  Research Assistant

  Latent Fingerprint Enhancement, Generative AI Project• Created a Generative AI system to enhance latent fingerprints, removing the background texture and obtaining a clear fingerprint structure free of anomalies.• Achieved SSIM 0.996, FSIM 0.947, UIQI 0.891, from GANs showcasing accuracy in enhanced fingerprints.• Engineered a NST to generate 500 realistic latent fingerprints, resolving the dataset limitation for model training.

• 

  Florida Institute for National Security (FINS)

  Aug 2023 - Dec 2023

  Research Assistant

  Computational Behavioural Analytics, Natural Language Processing Project• Collaborated with a team of 10 researchers to train a Natural Language Processing model to identify Big Five personality traits based on text, review, and tweets of the person for soft biometric identification.• Achieved a 77.8% F1 score in personality prediction through a comparative study and optimization of LLMs.• Reduced time by 98.73% using Hypergator with A100 GPUs for efficient LLM training and hyperparameter tuning. Show less

• 

  EDspare

  Jan 2024 - Aug 2024

  Software Development Engineer - Machine Learning

  Hyper Infinity Systems Quiz-2024, Computer Vision and LLM project ● Developed a system for Quiz platform to increase security and limit cheating by using state-of-the-art deep learning techniques for more than 10,000 users by incorporating Python, Machine Learning, AWS and AI.● Enhanced security and reliability by integrating Scikit-Learn and Deep Learning based proctoring system.● Reduced human error and bias by up to 65% through automated essay grading using GPT-based LLM models.● Engineered a nested RESTful API, achieving a 40% improvement in data retrieval speed for a user base of 100K+.● Enhanced user data security with JWT and bcrypt, significantly reducing unauthorized access.● Automated CI/CD processes with GitHub Actions, preparing for a Jenkins transition to boost scalability and reliability.● Implemented AWS ALB and EC2 auto-scaling, improving load distribution and resource efficiency by 25%.● Achieved a 20% reduction in operational costs while enhancing system scalability for an expanding user base Show less

• 

  ElevateMe

  Sept 2024 - Jan 2025

  Data Analytics & Machine Learning Fellow Trainee

  Completed 150+ hours of hands-on learning and project work, including a live project and two optional capstone projects.Gained practical experience applying data analytics and machine learning techniques to solve real-world problems.Worked with advanced tools such as Python, Jupyter Notebooks, Pandas, Numpy, Matplotlib, Seaborn, and Scikit-learn to build and evaluate ML models.Deployed machine learning models on Microsoft Azure, connecting Azure ML to Azure SQL databases for real-time analytics and predictions.Developed strong skills in data preprocessing, feature engineering, model selection, and evaluation through extensive use of regression, classification, and clustering algorithms.Collaborated in a team environment, simulating a professional setting with tools like Jira, Confluence, and Slack for project management and communication Show less

• 

  Amazon Web Services (AWS)

  Jan 2025 - now

  Software Development Engineer