Summer 2026 Technology Internship Program

Reduced data retrieval time by ~97% by developing a Python CLI using the AVEVA PI Web API, integrated into shell scripts for fully automated data ingestion.

Engineered a real-time AVEVA PI Vision monitoring system for high-volatility data center loads, enabling anticipatory generation dispatch to mitigate ACE swings.

Established the organization's best winter load-forecasting methodology using XGBoost, reducing winter-peak MAPE by 47% with meteorological inputs and load-growth analysis.

Built a BERT-based model to encode longitudinal patient histories into vector embeddings, to enable identification of undiagnosed patients, early risk flagging, and improved clinical trial cohort selection.

Performed delayed-entry Kaplan-Meier survival analysis using Python's lifelines package, delivering actionable mortality insights for patient risk assessment.

Conducted subpopulation analysis of CDKL5 Deficiency Disorder data via Komodo Health's Sentinel, validating anti-epileptic drug usage patterns to support clinical research.

Developed an AI analytics agent trained on client business semantics, letting executives query metrics in natural language and auto-generate dashboards, reducing analytics team backlog.

Improved agent response accuracy by 16% by creating a specialized natural-language-to-PostgreSQL training dataset, resolving forecast bias and accuracy metric queries.

Reduced AI infrastructure costs by 50% by transitioning to OpenAI's asynchronous Batch API, optimizing the embedding process for thousands of competitor product matches.

Standardized development environments using Mamba YAML files and documentation, reducing onboarding time and ensuring consistent project setups.