Summary
Overview
Work History
Education
Skills
Certification
Affiliations
Languages
Timeline
Projects
6e
Alyster Benedict

Alyster Benedict

HUBBALLI

Summary

AI/ML Engineer and Android Developer.

  • Final-year Computer Science student and academic topper at KLE Institute of Technology, with a consistent 8.8+ CGPA.
  • Skilled in developing and deploying end-to-end computer vision and AI models using Python, TensorFlow, and MediaPipe, with practical experience gained at the Samsung Innovation Campus.
  • Engineered 'Trainium,' a comprehensive AI fitness application for Android that provides real-time pose correction using computer vision, and personalized, gamified workout plans to enhance user engagement.
  • Experienced in software and game development using C++ (DSA) and Unreal Engine 5.

Overview

2
2
years of professional experience
1
1
Certificate

Work History

Intern

ASPL
Bengaluru
07.2025 - Current
  • Fine-tuning a Llama 2 7B parameter model on a specialized client dataset of ISO mechanical standards and documentation.
  • Implementing memory-efficient adaptation by leveraging the PEFT (Parameter-Efficient Fine-Tuning) library to apply QLoRA 4-bit quantization to reduce VRAM usage.
  • Currently focused on optimizing model performance and accuracy for, e.g., a specialized chatbot for engineers or a document analysis tool, by iterating on quantization levels and training parameters.
  • Collaborating with the team to preprocess and clean the technical dataset to ensure high-quality training input.

Student Intern

SAMSUNG INNOVATION CAMPUS
HUBBALLI
01.2024 - 11.2024
  • Engineered a real-time driver pose estimation system for in-car monitoring using Python, MediaPipe, and TensorFlow, achieving 95% accuracy.
  • Developed and implemented various machine learning models for real-world scenarios, managing the complete project lifecycle from data preprocessing, to model evaluation, and deployment.
  • Applied Agile methodologies to manage project timelines, and effectively collaborated with peers on a capstone project.
  • Presented regular project updates and technical findings during formal review meetings, demonstrating strong communication and presentation skills.

Technical Assistant

Codenex KLE INSTITUTE OF TECHNOLOGY
HUBBALLI
02.2025 - Current
  • Led training sessions on C++ and Unreal Engine for over for over 50+ students , covering core programming principles and advanced game development techniques.
  • Mentored student teams through hands-on projects, providing technical guidance and code reviews to solve real-time coding challenges.

Education

Bachelor of Engineering - Computer Science

KLE INSTITUTE OF TECHNOLOGY
HUBBALLI,KARNATAKA
01-2026

PUC - PCMCs

KLE PC JABIN SCIENCE COLLEGE
Hubballi,India
08-2022

High School -

St.Mary's High School
Hubballi,India
06-2020

Skills

    Domains & Core Competencies

  • Machine learning and AI
  • Computer Vision
  • Android Development
  • Unreal Engine Development
  • C Programming (DSA)

    Programming Languages

  • Python
  • C
  • Java
  • Kotlin
  • C

Certification

  • Samsung Innovation Campus AIML
  • Infosys AI course
  • Google Cloud AI
  • Gemini AI

Affiliations

  • Photography
  • Videography
  • Video Editing
  • Script Writing

Languages

English
First Language
Kannada
Proficient (C2)
C2
Tamil
Proficient (C2)
C2
Hindi
Proficient (C2)
C2

Timeline

Intern

ASPL
07.2025 - Current

Technical Assistant

Codenex KLE INSTITUTE OF TECHNOLOGY
02.2025 - Current

Student Intern

SAMSUNG INNOVATION CAMPUS
01.2024 - 11.2024

Bachelor of Engineering - Computer Science

KLE INSTITUTE OF TECHNOLOGY

PUC - PCMCs

KLE PC JABIN SCIENCE COLLEGE

High School -

St.Mary's High School

Projects

Trainium - AI-powered fitness trainer

  • Developed a novel, personalized workout generator that analyzes a user's photograph on-device. The feature uses a custom-trained, state-of-the-art EfficientNet model to estimate body metrics, achieving 95% accuracy in body type classification
  • Engineered a real-time pose correction engine using MediaPipe to track key body landmarks, providing live audio and visual feedback to ensure proper exercise form, and prevent injuries
  • Integrated a gamification system with experience points (XP), levels, and achievements to enhance user motivation and long-term engagement
  • Integrated a 24/7 conversational AI assistant to act as a fitness companion, providing users with motivational support and personalized advice
  • Technologies: Python, TensorFlow, OpenCV, Java/Kotlin, and Android

AI fitness tracker

  • Developed a Wear OS application to monitor and record user fitness metrics, like high heart rate anomaly detection, fall detection, etc., seamlessly syncing data with a companion Android phone app and a desktop dashboard for a unified user experience
  • Technologies: Android development, Java, Wear OS, Python, TensorFlow

Ray-traced RPG game

  • Developed a high-fidelity role-playing game (RPG) using Unreal Engine, implementing advanced real-time ray tracing to achieve realistic lighting and shadows
  • Scripted core game mechanics and character interactions using C++ and Blueprints
  • Technologies: Unreal Engine, C++
Alyster Benedict