Machine Learning Engineer with GenAI & MLOps Expertise

GOKUL RAVICHANDRAN

812-***-**** **********@*****.*** LinkedIn

Professional Summary

Machine Learning Engineer with 4+ years of experience developing and deploying production-grade AI and machine learning solutions across healthcare, cloud, and industrial automation domains. Specialized in Generative AI, NLP, Computer Vision, and LLMs, with expertise in building scalable, high-performance systems that bridge research innovation and real-world applications. Proficient in Python, TensorFlow, PyTorch, and LangChain, with strong hands-on experience in data engineering, MLOps, and model optimization. Adept at leveraging cloud platforms (AWS, Azure, GCP) to deliver robust, secure, and efficient AI applications, consistently improving accuracy, reliability, and business impact across enterprise environments. Technical Skills

Programming & Scripting: Python, R, SQL, Bash, JavaScript Machine Learning & Deep Learning: Scikit-learn, TensorFlow, PyTorch, XGBoost, LightGBM, LSTM, CNN, Transformers (BERT, GPT, T5, LLaMA)

Generative AI & LLMs: LangChain, LlamaIndex, GPT-4, Gemini, RAG Pipelines, Prompt Engineering, Reinforcement Learning

(PPO, DDPG), RLHF, Bandit Algorithms

Data Engineering & Processing: Apache Spark, PySpark, Snowflake, Apache NiFi, AWS Glue, Airflow, DVC, SQL, Pandas, NumPy, ChromaDB, FAISS, Pinecone

MLOps & Deployment: Docker, Kubernetes, FastAPI, MLflow, GitHub Actions, Azure Kubernetes Service (AKS), SageMaker, Vertex AI, CI/CD, Model Versioning, Monitoring

Computer Vision & NLP: YOLOv5/v8, OpenCV, FastAI, spaCy, NLTK, OCR, Vision-Language Models, Text Summarization, Sentiment Analysis

Cloud Platforms: AWS (S3, Lambda, SageMaker), Azure

(Cognitive Search, Databricks, AKS), GCP (Vertex AI, BigQuery) Databases & Storage: Snowflake, Redshift, MongoDB, PostgreSQL, Elasticsearch, ChromaDB

Visualization & Analytics: Power BI, Tableau, Grafana, Streamlit, Gradio

Tools & Dev Practices: Git, JIRA, Agile, Weights & Biases (W&B), Grad-CAM, SHAP, QGIS, EconML, DoWhy

Professional Experience

AI/ML Engineer Cardinal Health, USA Feb 2025 - Present

Developed a GenAI-powered clinical assistant (ReX-Health) using GPT-4, Gemini, and LangChain, integrating RAG, Pinecone, and LlamaIndex to improve clinical query accuracy by 35% and increase physician adoption by 30%.

Engineered scalable vector search pipelines with FAISS, ChromaDB, and OpenAI embeddings, enabling sub-second retrieval of EHR and medical knowledge base records across 50K+ patient documents.

Built automated CI/CD workflows using Docker, MLflow, and GitHub Actions, reducing model deployment time from hours to minutes and ensuring audit-ready healthcare compliance.

Containerized multi-LLM microservices via FastAPI and Kubernetes, supporting autoscaling clinical inference endpoints across Azure Kubernetes Service (AKS) for large-scale healthcare workloads.

Deployed an end-to-end model monitoring system using Prometheus and Grafana, tracking model drift on diagnostic models and reducing clinical prediction errors by 22%.

Integrated Azure Cognitive Search for hybrid semantic + keyword retrieval of drug, procedure, and patient literature, boosting clinical query accuracy by 41%.

Designed RLHF-style feedback loops using clinician validation, prompt ranking, and Bandit algorithms, enhancing model reliability and improving AI-driven decision support for healthcare professionals. Machine Learning Engineer Hellinex Cloud, India Dec 2020 - Jul 2023

Developed digital twin simulation models using Reinforcement Learning (PPO, DDPG) in PyTorch, optimizing irrigation and fertilizer scheduling strategies that increased simulated crop yield accuracy by 18%.

Built real-time soil anomaly detection systems utilizing YOLOv5, OpenCV, and geospatial QGIS overlays, delivering 92%+ detection precision and enabling data-driven agronomic intervention across monitored regions.

Engineered scalable data ingestion pipelines using Snowflake and Apache Spark to automate the processing of satellite and IoT sensor data, enabling faster analysis and improving the efficiency of multi-source data workflows.

Researched and authored an internal whitepaper on Causal Inference in Deep Learning employing DoWhy, EconML, and SHAP, enhancing model explainability and stakeholder confidence in ML-driven predictions.

Delivered interactive Power BI dashboards integrated with prescriptive AI insights, empowering cross-functional teams to make real-time data-backed decisions and improving operational response efficiency by 25%. Machine Learning Engineer Parallax Pros, India Jun 2019 - Nov 2020

Developed and deployed CNN-powered visual defect detection systems using TensorFlow and TFLite on edge devices, enhancing industrial IoT quality inspection throughput by 40% across automated production lines.

Implemented a real-time fraud detection engine leveraging Isolation Forests, Autoencoders, and Z-score analytics, reducing false positives by 33% and improving transactional reliability across enterprise systems.

Built an LSTM-based predictive maintenance framework on Azure Databricks, forecasting equipment failures with 28% higher accuracy, minimizing unplanned downtime, and optimizing maintenance scheduling.

Automated complex ETL workflows using Apache Airflow and DVC, cutting data pipeline latency by 45% and enabling scalable, version-controlled model retraining across distributed data environments.

Optimized data storage and retrieval in MongoDB and PostgreSQL via vector-based similarity search, improving dashboard response time from 5 seconds to under 1.2 seconds for real-time analytics. Education

Masters of Science in Computer Science Aug 2023 - May 2025 Indiana State University, Terre Haute, IN

Bachelors of Engineering in Computer Science Jun 2016 - Aug 2020 Anna University, Chennai, India

Projects

SmartCrop Vision – Precision Agriculture Image Diagnosis

Developed a precision agriculture system using CNNs, FastAI, FastAPI, and Docker to identify crop diseases from field images; integrated MongoDB and Power BI dashboards for real-time visualization, delivering 92% classification accuracy and reducing manual inspection time by 40% through Agile sprints in JIRA. NewsSense – Real-Time News Summarization and Sentiment Pipeline

Engineered a real-time NLP workflow using T5, BART, and Pegasus from Hugging Face Transformers, combined with spaCy, NLTK, and custom regex for text preprocessing; deployed via FastAPI and Kafka to stream news summaries and sentiment alerts, reducing information processing time by 60% for research teams.

Contact this candidate