Allu Likhith

Likhith Allu- - Mudinepalli, Andhra Pradesh Professional Summary Bachelor's in Computer Science Engineering and expertise in Java and Python. Proven ability to design, develop, and optimize scalable software solutions on AWS. Adept at leveraging data structures, algorithms, and cloud technologies to drive product improvements and efficiency gains. Collaborative team player focused on delivering high-quality, impactful results through continuous learning and modern development practices. Experience Blockchain Developer Metaacrafters | Remote 2023/JUN – 2023/SEP Reduced gas consumption in existing smart contracts by 15%, saving 1000+ users $5,000 in transaction fees over 3 months and improving platform performance. Developed and deployed a decentralized identity solution using Hyperledger Fabric for a consortium of 5 partner organizations, decreasing identity verification costs by 30% and enhancing data security for 5000+ users. Projects 2025/JAN – 2025/MAY Genarative AI for Creating Synthetic Data from Videos Designed and implemented a GAN-based pipeline using TensorFlow and Python, increasing object detection model accuracy by 30% compared to training on solely real-world data, impacting 10,000+ model iterations. Optimized GAN training process by implementing a custom loss function and hyperparameter tuning, achieving a 20% reduction in training time and a 15% improvement in synthetic image fidelity (FID score). 2024/JUN – 2024/DEC Object Detection Developed a YOLOv5 object detection pipeline using PyTorch, enhancing detection accuracy by 15% on a dataset of 10,000+ images through optimized hyperparameter tuning and custom loss function implementation. Imbalance issues using a focal loss function and data augmentation techniques, resulting in a 20% improvement in F1-score for minority classes during object detection, enabling more accurate event classification. 2023/SEP – 2024/JAN Sentimental Analysis for Product Review Built a sentiment analysis model using BERT and TensorFlow, improving accuracy by 15% over existing solutions in classifying product reviews across a dataset of 1000+ entries, improving customer insights. Implemented a scalable data pipeline with Apache Kafka and Spark to process real-time product reviews, reducing latency by 40% and enabling immediate identification of negative feedback trends within 15 minutes. Refined the sentiment analysis algorithm using hyperparameter tuning and feature engineering, leading to a 20% reduction in processing time while maintaining accuracy and decreasing computational costs by 10%. Certifications 2025/JUN AI Engineer Oneroadmap 2024/JUL Introduction to Cybersecurity Cisco 2023/JUN Blockchain Technology using Ethereum and Polygon Metacrfters Skills Java Python AWS Cloud SQL Data Structures and Algorithms JavaScript TensorFlow PyTorch API Development Scalable Systems Education Chandigarh University Bachelor's of Engineering in Computer Science 2021/08 – 2025/05