Sai Nikhil Thirandas

Lead Machine Learning Engineer at AWS Bedrock with 11 years of experience across distributed systems, machine learning, and generative AI. Architected Helios, an LLM/Agent evaluation framework. Holds a B.Tech (Hons.) from IIT Kharagpur and M.S. in Applied Mathematics from Northeastern University. Strong foundation in Data Structures & Algorithms, System Designing, and Mathematics (Calculus, Linear Algebra, Probability & Statistics). Core expertise includes LLM/Agent evaluation, ML Model Serving, Computer Vision, NLP, and Distributed Computing. Built quality classifier models and distributed inference pipelines. Designed custom BPE tokenizer (256K vocab) for Nova models. Evaluated LLMs on Artificial Analysis Intelligence Index benchmarks. Proficient in PyTorch, TensorFlow, LangChain, CrewAI, and Google ADK for building production ML systems. Co-authored research paper on domain-specific pre-training (DoPAMine). Passionate about designing innovative solutions that improve accuracy, efficiency, and reliability in production ML systems.

Python, Java, R, C/C++, MATLAB, Mathematica, SQL, PHP, Perl, HTML, CSS, TypeScript, XML, JSON, Visual Basic

LLMs, Generative AI, Agent Systems, LLM/Agent Evaluation, Computer Vision, NLP, Deep Learning, Regression, Classification, Ranking, Recommendation Systems, Clustering, Dimensionality Reduction, Feature Engineering, OCR

PyTorch, TensorFlow, DeepSpeed, vLLM, HuggingFace Transformers, LangChain, CrewAI, Google ADK, OpenCV, NumPy, pandas, Matplotlib, scikit-learn, SymPy, Spark, Hadoop

AWS SageMaker, AWS EMR, AWS Lambda, AWS S3, AWS Step Functions, Git, Docker, Kafka, Hive, Zookeeper, Elasticsearch, Linux, Jupyter Notebook, PyCharm, IntelliJ IDEA

This article contains verified ✅ certification links to all the courses I completed on Coursera.