AI/ML Engineer

• AI/ML Engineer with 9+ years of expertise in Data Science, Machine Learning, Deep Learning, and Large Language Models (LLMs), specializing in building, fine-tuning, and deploying AI-driven solutions at scale.
• Proficient in Python and experienced in end-to-end ML workflows, including data collection, preprocessing, EDA, model training, and deployment using cloud and MLOps practices.
• Expertise in Supervised and Unsupervised Learning algorithms for classification, regression, clustering, and anomaly detection.
• Designed machine learning algorithms such as XGBoost, Random Forest, Support Vector Machines (SVM), Logistic Regression, and K-Means, implementing techniques like outlier detection, hyperparameter tuning, cross-validation, and feature engineering, resulting in up to a 15% improvement in model accuracy.
• Hands-on experience with Deep Learning frameworks such as TensorFlow, PyTorch, and Keras, implementing CNNs, RNNs, LSTMs, GANs, and transformer-based NLP models for various AI applications.
• Fine-tuned and deployed LLMs such as GPT, BERT, LLaMA, and Falcon for NLP applications like text generation, summarization, and Q&A.
• Implemented Retrieval-Augmented Generation (RAG) architectures by integrating LLMs with Vector Databases (FAISS, ChromaDB, Weaviate, Pinecone) to enhance search and response accuracy.
• Optimized LLM inference using quantization techniques (GPTQ, LoRA, DeepSpeed, and Hugging Face Transformers) to reduce latency and improve efficiency.
• Processed large-scale data using HDFS, MapReduce, and PySpark, optimizing model training for high-performance and distributed ML workloads.
• Designed and implemented scalable ML pipelines leveraging big data technologies to handle structured and unstructured datasets.
• Built and deployed AI models using FastAPI, Flask, and Streamlit, ensuring real-time inference and seamless integration with applications.
• Containerized AI models with Docker and Kubernetes, deploying them on cloud-based infrastructures for scalability and high availability.
• Developed serverless and microservices-based ML architectures for efficient model inference and API exposure.
• Worked with AWS, GCP, and Azure, utilizing AI/ML services such as SageMaker, Bedrock, Vertex AI, AutoML, Azure ML & AI, EC2, Lambda, BigQuery, and S3 for scalable model training and deployment.
• Designed CI/CD pipelines using GitHub Actions, MLflow for model tracking, and Terraform for cloud infrastructure automation.
• Implemented monitoring, logging, and performance tracking using AWS CloudWatch, GCP Cloud Monitoring, and Azure Monitor for observability.
• Integrated security and audit logging using AWS CloudTrail logs to track model performance, API access, and system events in production.
• Implemented monitoring, logging, and performance tracking with Prometheus, Grafana, and ELK stack for AI models in production.
• Deployed containerized workloads on Kubernetes clusters (EKS, GKE, AKS), managing cloud-based MLOps pipelines for automated retraining and deployment.
• Proficient in evaluation metrics such as precision, recall, F1-score, and AUC-ROC, ensuring 95%+ accuracy and 90% precision in classification and regression tasks.