Kevin Powell

Senior ML Engineer

Surescripts

Jun 2023 - Current (3 years 1 month)

• Promoted from Mid-Senior ML Engineer after 9 months of high-impact delivery. • Led development and deployment of machine learning models to accelerate drug discovery and enhance clinical trial analysis, reducing R&D timelines by ~20%. • Designed and implemented LLM-powered internal chatbots using LangChain and OpenAI GPT APIs, boosting researcher productivity by 35% via faster literature insights and structured data access. • Built generative AI systems for automated data synthesis, enabling rapid generation of trial summaries and internal reports—cutting manual effort by 50%. • Developed end-to-end NLP pipelines for extracting and analyzing clinical notes, patient records, and scientific publications, reducing abstraction time from

Mid Senior ML Engineer

Surescripts

Sep 2022 - May 2023 (9 months)

• Contributed to backend ML infrastructure supporting friend recommendation systems, resulting in a 5– 10% improvement in user engagement. • Designed and optimized large-scale recommendation algorithms to balance inference latency and predictive accuracy across 200M+ daily active users. • Developed internal tools for LLM-based content moderation, reducing latency by ~25% and enhancing precision for detecting nuanced, policy-sensitive content. • Researched and applied Graph Neural Networks (GNNs) to improve social graph modeling, boosting the quality of friend suggestions through richer relationship representations. • Built and maintained prompt engineering libraries adopted by 4+ internal teams to standardize LLM use across moderation,

ML Engineer

Omnicell

Apr 2018 - Aug 2022 (4 years 5 months)

• Built and deployed predictive models using XGBoost, LightGBM, and Random Forests, increasing forecasting accuracy by 20–30% across claims prediction, patient risk scoring, and churn analysis. • Designed scalable NLP pipelines using spaCy, Hugging Face Transformers, and Gensim to process and classify 1M+ documents for use cases such as clinical NER, claims triage, and contract analysis. • Developed early-stage NLP tools with spaCy, BERT, and TextBlob for entity extraction, keyword tagging, and summarization — reducing document review time by ~50% for pilot teams. • Engineered and launched a hybrid rule-based/ML chatbot using Rasa, which reduced average internal IT ticket resolution time by 40% and improved SLA compliance; later extende

Software Engineer

iDonate

May 2016 - Mar 2018 (1 year 11 months)

• Built and deployed robust Python-based backend services, improving system reliability and reducing runtime errors by ~40% in production environments. • Automated complex data processing workflows, reducing manual effort by 60% and improving reporting accuracy by 25% across internal BI systems. • Owned end-to-end delivery of 10+ client-facing projects, gathering requirements and working cross functionally to consistently exceed delivery timelines and quality expectations. • Wrote internal documentation and led onboarding sessions, accelerating ramp-up time for new developers by 30% and improving team codebase fluency.

About

Detail

Timeline

Résumé