Karthik Viswanathan S. Email & Phone Number

Santa Clara, Ca, Us

Sunnyvale, California, Us

As a Solutions Architect, I help in enabling customers iterate an order of magnitude faster on research and development of Generative AI applications by leveraging hardware and software optimizations of the Cerebras compute platform.Deep technical expertise in helping customers understand the fundamentals of their distributed training bottlenecks in their Generative AI journey, with a specific focus on data center infrastructure coupled with software scaling optimizations that enable pre-training and fine-tuning of GPT style models leveraging domain specific textual data corpuses.

Cambridge, Cambridgeshire, Gb

Consultant data scientist on AstraZeneca’s digital health initiative, with the aim of enhancing clinical trial workflows with human-in-the-loop AI.Developed novel clinical workflows applying hybrid LLM+Classical ML approaches for Named Entity Recognition (NER), Relation Extraction (RE) and classification of Clinical Event Endpoints under Adjudication by medical experts.The system can automate identification commonly occurring outcomes of endpoints in Cardiovascular Clinical trials, reducing time to market for new drugs/vaccines, triaging only complex events for manual review.

Durham, North Carolina, Us

As a ML Engineer at IQVIA, I developed data science models focusing on automating key aspects of clinical trials workflows and processes such as adverse event adjudication, pre-screening of prospective patients for clinical trial recruitment and screening of at risk patients for rare diseases, helping save time through reduction of manual effort, which helps in bringing drugs to market faster, and value through capability of scaling systems to serve the need of more patients.This included development of both traditional machine learning approaches for critical production applications, such as tree based models, bayesian classifiers for resource constrained problems that needed high degree of precision and explainability in a clinical setting.Worked with research scientists to scale deep learning model training for drug discovery research problems like DeepPurpose.Leveraged the applied ML expertise in clinical trials domain as a data science consultant for AstraZeneca, helping prototype data science based workflow enabling human-in-the-loop AI assisted automation of adverse event adjudication in cardiovascular clinical trials.

Milford, Ma, Us

Investigate development of deep learning techniques for classification of proteomic data generated from biological entities using Liquid Chromatography and Mass Spectrometry. Develop machine learning pipeline for general end-to-end application on such classification tasks.Technologies used:Tensorflow, Keras, Python

Boston, Us

Teaching Assistant for graduate Computer Systems course CS5600 which teaches advanced OS concepts for distributed computing.Technologies used:C, Golang

Design and development of hardware for telecommunication networks with emphasis on wired network infrastructure at the exchange, router and gateway. Handling hardware support services for legacy TDM networks.

Master'S Degree, Computer Science

Bachelor'S Degree, Electrical, Electronics And Communications Engineering

High School, Computer Science

High School