Built LUNA, automating 20% of daily satellite tickets and cutting resolution time by 90% by replacing manual analyst triage with a decision tree engine and fault-tolerant retry logic across Databricks and EKS.

Extended LUNA to apply tuning templates across LeoLabs' entire satellite catalog (the world's largest) by replacing a workflow bottlenecked on 7 AWS Workspaces licenses with an SQS-queued pipeline running across 20 parallel EKS licenses.

Designed a predictive test orchestration engine that analyzes dependency trees to run only affected tests, cutting sandbox environment CI/CD monthly compute spend by 18% (roughly $6,500/mo).

Repackaged two shared modules for distribution via CodeArtifact, eliminating subprocess-based cross-repo dependencies across 5 satellite tracking cron jobs and enabling stable software releases through version control.

Retargeting campaigns were reaching only a fraction of the real user base; engaged app users simply weren't in HubSpot. Built an event-driven sync between the application and HubSpot Contacts, with deduplication logic to prevent conflicting records across data sources, and exposed usage properties so marketing could segment audiences by real behavior. This grew the matchable retargeting audience by 62%.

Built a RAG system for student exam prep, but the interesting engineering problem was evaluation, not retrieval or generation. Designed the pipeline around retrieval-confidence scoring and citation gating, then built an adversarial test set using unrelated MCAT sections as distractor content to measure whether the system knew when to abstain rather than hallucinate. It achieved 82% correctness on answerable queries and full abstention on every adversarial input, driving adoption by roughly 40% of users during peak test periods.

Architected the REST APIs and Postgres schemas behind cohort reporting and risk segmentation from scratch. Once in production, table loads slowed significantly under concurrent access; diagnosed the bottleneck as expensive join and count queries on unindexed columns, then added targeted indexes and a Redis caching layer storing precomputed cohort profiles, cutting page loads from 5 seconds to 1 second.

Built the data pipeline behind the Cohort Risk dashboard, aggregating multi-source engagement signals (login frequency, assignment progression) into predictive risk indicators surfaced directly to instructors. This drove a 16% lift in daily active instructor usage of the tool.

Engineered temporal features from GOES satellite solar flux data (lag windows and flux rate-of-change) and designed time-series cross-validation splits specifically to eliminate lookahead bias, a common and easy-to-miss failure mode when forecasting rare, irregular events like SPEs. This reduced feature importance variance by 11% across folds, giving a more reliable signal to build on.

Built the end-to-end modeling pipeline comparing a LightGBM baseline against a Temporal Convolutional Network, using nested cross-validation for hyperparameter selection. The final ensemble outperformed NASA's existing prototype by 6% on out-of-distribution solar event windows, the evaluation setting that actually matters, since SPEs are exactly the kind of rare event a model needs to generalize to, not memorize.

Refactored all 12 legacy PII endpoints with JWT authentication and AWS KMS-backed encryption middleware, unblocking UNIPORT's B2G to B2B transition by meeting enterprise-grade security requirements.

Consolidated PII endpoint access to 2 authorized services via granular IAM controls, reducing blast radius by 83% and limiting data exposure to strictly necessary service principals.