I’m a Master’s student in Computer Science at Cal State Fullerton, focused on building AI and machine learning solutions that solve real-world problems. My experience spans developing scalable MLOps systems, computer vision research, and automated AI pipelines.

At VSQUARE, I improved eye treatment diagnostics by 15% through computer vision models that integrated emotion recognition with 92% accuracy. I’ve built complex systems like DevOps Orchestra, which automates 95% of software delivery workflows using AI agents and Kafka, reducing deployment times and human errors.

Other projects include a real-time credit card fraud detection pipeline on AWS Lambda with sub-3-second latency, leveraging MLOps tools like DVC, Hydra, Weights & Biases, Evidently AI, and MLflow to ensure scalable, reliable, and monitored model deployment. I also developed a fact-checking news aggregator that processes 50,000+ articles daily with improved accuracy. My research in neural style transfer advanced image quality by 20% through novel transformer-based techniques.

I hold AWS Certified AI Practitioner and AWS Certified Cloud Practitioner certifications, which complement my technical skills by validating my expertise in designing and deploying AI solutions on the AWS cloud. These certifications have strengthened my ability to build scalable, secure, and cost-effective cloud-native AI systems that integrate seamlessly with modern infrastructure.

I work primarily with Python, PyTorch, TensorFlow, AWS, Kafka, Docker, Terraform, and MLOps frameworks to combine deep learning with robust infrastructure and deliver measurable impact.

• Devops-orchestra
  AI-powered DevOps automation framework built on a collaborative multi-agent architecture, where agents handle the entire SDLC—from code validation to cloud provisioning, testing, deployment, and recovery—triggered by GitHub and orchestrated via Kafka.

• Real-Time Fraud Detection Pipeline
  Production-ready pipeline with PyTorch Lightning, ONNX, AWS Lambda, MLOps tools (DVC, MLflow, W&B, Evidently AI), and automated CI/CD.

• Real-Time Election Voting Analysis
  Election trend forecasting with ARIMA and linear regression, secure voter verification using Siamese networks, Kafka streaming, Hadoop storage, and Streamlit visualizations.

• Neural Style Transfer with Transformers
  Transformer-based style transfer enhanced by pyramidal positional encoding and reinforcement learning, boosting image quality and training efficiency.

• DATON - Vehicle Number Plate Detection & Tracking
  Real-time vehicle tracking system using YOLOv8, WPOD-net, and EasyOCR, enhancing detection accuracy and processing speed.

• Email: [email protected]
• LinkedIn: linkedin.com/in/raahulkrishna
• Portfolio: raahul-tech.github.io/Raahul_Portfolio_Website

Outside coding, I mentor peers, lead AI workshops, and design explainable and scalable AI systems.