Hi there, I'm

Sri Ganesh Arjula

Machine Learning and Generative AI Engineer specializing in deploying real-time predictive models, NLP solutions, and scalable AI systems. Expert in LLM fine-tuning, cloud platforms, and optimizing business operations with AI-driven solutions.

View Projects View Resume Get in Touch View My Models

< About Me />

I am Sri Ganesh Arjula, a passionate Machine Learning and AI Engineer with a knack for developing innovative solutions that drive business transformation. With a strong foundation in deploying real-time predictive models and NLP solutions, I excel in creating scalable AI systems that enhance operational efficiency. My expertise spans across various industries, including logistics, healthcare, and banking, where I have been recognized for my contributions to AI and IoT projects. I am committed to leveraging AI to solve complex problems and am always eager to explore new technologies and methodologies.

Traditional ML Algorithms
Scikit-learn, XGBoost, LightGBM, Random Forests, SVM
Deep Learning
TensorFlow, PyTorch, Keras, CNNs, RNNs
Computer Vision
OpenCV, YOLO, Object Detection, Image Segmentation
MLOps & Deployment
MLflow, Docker, Kubernetes, FastAPI, Model Monitoring
Cloud ML Services
AWS SageMaker, Azure ML, Google Vertex AI
Big Data Processing
Spark, Hadoop, Kafka, Data Pipelines
Large Language Models
LLaMA, GPT, BERT, Transformers, Fine-tuning
Gen AI Frameworks
LangChain, Hugging Face, OpenAI API, Anthropic API
Prompt Engineering
RAG, Few-shot Learning, Chain of Thought, System Prompts
Gen AI Infrastructure
vLLM, Text Generation Inference, Model Quantization
Multimodal Gen AI
Stable Diffusion, DALL-E, GPT-4V, Image Generation
Gen AI Tools
LlamaIndex, Semantic Kernel, Vector Databases

< Work Experience />

Machine Learning / Generative AI Engineer

ArcBest Corporation, Fort Smith, AR

Jul 2024 – Present
  • ▹ Developed an AI-driven auditing system to automate financial and logistics document processing, reducing manual workload by 40% and improving fraud detection accuracy by 35%.
  • ▹ Fine-tuned LLaMA using LoRA and PEFT, optimizing document auditing, compliance validation, and fraud detection insights.
  • ▹ Implemented AI-driven OCR and document validation, improving structured data extraction accuracy from invoices, contracts, and shipping documents to 95%.
  • ▹ Designed an AI-powered compliance validation workflow, generating automated compliance summaries and fraud insights from financial transactions and shipment records.
PyTorch AWS Azure ML Kubernetes

Machine Learning Engineer

Karnataka Bank, Bangalore, India

Apr 2021 – Jun 2022
  • ▹ Developed an enterprise-scale Credit Risk Prediction system using ensemble models, improving accuracy by 25% and significantly reducing loan default rates.
  • ▹ Designed and implemented advanced Feature Engineering pipelines, incorporating behavioral scoring and credit utilization patterns, increasing model precision and lift by 35%.
  • ▹ Developed containerized model deployment solutions and RESTful APIs, enabling real-time credit scoring with 99.9% system uptime.
  • ▹ Established automated CI/CD pipelines, improving model deployment frequency by 50% and ensuring seamless integration into operational processes.
TensorFlow PyTorch Hadoop Apache Spark Kafka

Machine Learning Engineer

Liberty Insurance, Bangalore, India

May 2020 – Mar 2021
  • ▹ Developed and deployed Time-Series Forecasting models using ARIMA, SARIMA, and Prophet to predict claim frequency and customer acquisition trends.
  • ▹ Built real-time data pipelines for fraud detection and claims processing, ensuring low-latency data streaming and efficient data handling.
  • ▹ Designed and containerized machine learning models using Flask and Docker, ensuring scalable and high-availability deployments.
  • ▹ Optimized data integration and real-time analytics, enabling faster query processing and improved decision-making for business intelligence teams.
ARIMA SARIMA Prophet Kafka Spark

< Featured Projects />

View Code

Skin Care AI Chatbot

Advanced dermatology chatbot leveraging OpenBioLLM and Pali-Gemma for accurate skin disease diagnosis and personalized treatment recommendations. Features real-time image analysis and medical knowledge graph integration.

Python FastAPI OpenBioLLM Pali-Gemma PyTorch
Live Demo

P2P Community Platform

Developed a FastAPI-based backend for ridesharing, service requests, and rental listings with real-time WebSockets for notifications.Implemented advanced search and routing functions, enabling users to combine multiple services and manage accounts via IVR using Twilio and groq.Designed a dynamic, responsive interface using HTML, CSS, and JavaScript for seamless user interactions

Python FastAPI React MongoDB Google Maps API
View Code

RAG Based Chatbot

Intelligent webpage chatbot utilizing Retrieval-Augmented Generation with ChromaDB for efficient document querying. Features semantic search, context-aware responses, and real-time information retrieval from dynamic content.

Python FastAPI ChromaDB LangChain Llama 2
View Code

Heart Sound Classification

Advanced deep learning system using RNN architecture to classify heart sounds into normal, murmur, and artifact categories. Implements signal processing techniques and achieves 95% accuracy in real-world clinical settings.

Python TensorFlow Keras Librosa NumPy

Ferrell Gas Service Center Clustering

Interactive analytics dashboard with real-time performance metrics visualization. Implements K-Means clustering for service center categorization and features an AI chatbot for data insights using BERT-based models.

Python Dash Plotly scikit-learn BERT PostgreSQL

American Sign Language Recognition

Developed an ASL recognition system using deep learning. Collected and labeled the ASL Alphabet Dataset from Kaggle, adding bounding box annotations with LabelImg. Used transfer learning with ResNet and MobileNet, achieving 92% and 86.76% accuracy, respectively, with GPU acceleration (RTX 2060). Implemented using TensorFlow. The system supports static ASL signs but does not detect letters ‘J’ and ‘Z’ due to motion-based gestures.

Python TensorFlow LabelImg ResNet MobileNet GPU Acceleration

< Education & Achievements />

Education

Master of Science in Data Science

Wichita State University, Wichita, KS

May 2024 | GPA: 3.8/4

Certifications

Awards

Publications