About
Highly accomplished AI/ML Engineer with over 2 years of experience in designing, optimizing, and deploying production-ready AI/ML models across diverse platforms including CoreML, ONNX, and TensorRT. Proven expertise spans computer vision, deep learning, and large language models, evidenced by a 40% improvement in training efficiency and 65% reduction in inference costs while consistently maintaining 95%+ accuracy. Adept at developing scalable ML backends and real-time AI decision systems for mission-critical applications.
Work
→
Summary
Provided expert AI/ML engineering services, specializing in model fine-tuning, NLP pipeline development, and cross-platform AI solution deployment for diverse clients.
Highlights
Fine-tuned 20+ deep learning models (GPT-4, Llama, Falcon) using PyTorch and TensorFlow, achieving 90% accuracy on domain-specific tasks while reducing inference costs by 50%.
Implemented advanced NLP pipelines with transformer architectures for entity extraction, sentiment analysis, and intent classification, maintaining 98% uptime for mission-critical business operations.
Developed cross-platform AI solutions deployed on Mac, Windows, and mobile platforms, optimizing models using ONNX and CoreML for edge deployment with 65% cost reduction.
Built computer vision systems processing 1000+ daily images with 92% accuracy, implementing CNN architectures with custom data augmentation and real-time inference optimization.
→
Summary
Led the design and implementation of advanced AI solutions, focusing on LLM fine-tuning and scalable ML backend development for production environments.
Highlights
Designed and implemented a localized fine-tuning pipeline for LLMs using PyTorch, boosting training efficiency by 40% and enabling secure on-premises model development.
Integrated Unsloth optimization framework to enhance GPU usage efficiency by 4x, enabling 100x faster LORA training and supporting 7B parameter model fine-tuning on limited hardware resources.
Developed autonomous AI agents with TensorFlow-based models for market research, achieving 85% accuracy in product potential analysis and customer sentiment classification.
Built a scalable ML backend using FastAPI with optimized model serving infrastructure, implementing professional MLOps practices for production deployment.
→
Summary
Conducted research and development in real-time AI decision systems and optimized deep learning model deployment for edge devices.
Highlights
Designed real-time AI decision systems using TensorFlow, reducing model inference latency from 500ms to 75ms through advanced optimization techniques and hardware acceleration.
Implemented comprehensive model monitoring frameworks with automated QA protocols, achieving 99.9% system reliability for production AI deployments.
Optimized deep learning model deployment pipelines for edge devices using TensorRT and quantization techniques, reducing inference costs by 65% while maintaining 95% accuracy.
Developed a parking space detection system using convolutional neural networks for LiDAR scan analysis, implementing high-precision real-time classification.
Awards
First Prize, EMU INVENT 2024
Awarded By
EMU INVENT
Awarded for AI-integrated Brain-Controlled Wheelchair using on-device ML models.
MAAJEE Scholarship Recipient 2023
Awarded By
MAAJEE
Awarded for academic excellence in AI/ML coursework.
Languages
English
Skills
Platforms
Mac, Windows, Linux, Mobile (iOS/Android), Edge Devices.
Tools & Libraries
HuggingFace, Docker, Git, Jupyter, FastAPI.
AI/ML Frameworks
TensorFlow, PyTorch, Keras, Scikit-learn, OpenCV.
Model Optimization
ONNX Runtime, TensorRT, CoreML, Quantization, Pruning, Unsloth, PEFT, LoRA.
Deep Learning
CNNs, RNNs, Transformers, YOLO, U-Net, ResNet, Attention Mechanisms, LLMs, Fine-tuning.
Computer Vision
Object Detection, Segmentation, Tracking, Feature Extraction, LiDAR Scan Analysis, Data Augmentation.
Programming Languages
Python, C++, CUDA, JavaScript, SQL.
Cloud & Deployment
AWS, GCP, Azure, MLOps, CI/CD, Model Monitoring, Production Deployment, Hardware Acceleration.
Problem Solving
System Reliability, Performance Optimization, Real-time Systems.