Anirudh Balakrishnan

I’m an Applied AI Engineer with experience designing and deploying production machine-learning systems, agentic AI applications, and cloud-native data platforms. At Caterpillar, I delivered large-scale anomaly-detection solutions for telematics data and enterprise MCP-based AI systems, engineered for real-world reliability and impact.

I built a production-grade anomaly-detection pipeline for real-time telematics streams from 1,000+ heavy-equipment machines across 25 product groups, combining HDBSCAN for unsupervised outlier detection with XGBoost classification to achieve an 80% accuracy improvement over rule-based systems. During my intern work at Caterpillar, I designed Snowflake automation workflows to orchestrate cross-region replication of 3,000+ tables and added real-time observability and alerting for proactive failure detection.

I’ve also shipped customer-facing mobile features as a Software Developer Intern across two fintech platforms, integrating CRIF credit-report APIs for real-time credit scoring and building reusable Flutter UI components for end-to-end workflows. My projects include an agentic RAG news digest with LangGraph and Qdrant, an AI infrastructure generation agent for VS Code Copilot Chat using MCP-based orchestration, a cricket analytics app (Flutter/Flask) with fine-tuned Hugging Face models, and time-series research using statsmodels, pandas, and SciPy.