Christopher Johnson

About Christopher Johnson

I have delivered more than 30 successful projects in the data science and machine learning areas. I have worked with both classical and deep learning solutions in a few industries. Also I am focused on creating systems regarding MLOps best practices and design patterns. Having experience with many cloud providers, I am able to automate the whole ML process from data gathering to automated deployments and continuous training.

Experience

1. Senior Ai Ml Engineer

   Nike

   May 2023 — Present

   Using Agile methodology to develop a project while working on a team. Developed a Bayesian inference and GPT-4 based fact-checking algorithm, significantly improving educational content verification accuracy. Created 15+ GPT-4 prompts, boosting AI interactions and content generation on our platform. Building AI prototype, establishing 22 Lambda apps, 13 DynamoDB tables, and 24 AppSync queries for a scalable NLP application. Enhanced conversational AI by fine-tuning Mistral model, integrating GPT-4 with a RAG system for a 40% increase in chatbot performance. Expertly deployed scalable NLP and Conversational AI solutions using AWS (DynamoDB, SageMaker, SAM, Cognito, Lambda, SNS, S3, Secrets Manager, AppSync, CloudFormation), cutting operational costs. Participated in features engineering such as feature generating, PCA, feature normalization and label encoding with Scikit-learn pre-processing. Work with NLTK library to NLP data processing and finding the patterns. combined LLM, deep learning, statistical NLP, syntactic, and lexical approaches to real-world NLP problems wrote production code for high-precision pattern matching based on linguistic features implemented feature extraction and data pre- and post-processing for NLP models engineered LLM prompts for production features

Education

• The University of Texas at Dallas

  Bachelor's degree

  2010 — 2014