Scholar
Aerin Young Kim
Google Scholar ID: Cj9C_d4AAAAJ
Miraflow AI, TrueMedia.org
Citations & Impact
Citations
485
H-index
8
i10-index
8
Publications
16
Co-authors
15
list available
Contact
No contact links provided.
Publications
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Research Experience
- Focused on building and optimizing large-scale production-grade machine learning models.
- ML Applications Development: Semantic parsing (text-to-SQL, QA, machine translation), deepfake detection/generation, visual QA, multimodal summarization, cross-modal retrieval, chatbots, object detection, semantic segmentation, and targeted ad delivery.
- Search Accuracy and Database Integrity: Enhanced search fidelity via advanced deduplication and indexing, improving operational efficiency and reducing costs.
- Training Dataset Synthesis: Created datasets for academia and industry using synthetic and human-in-the-loop methods.
- Stochastic Modeling and Simulation: Applied probability theory (Markov chains, queueing theory, reliability analysis) to simulate real-world phenomena.
- High-Volume Data Processing: Implemented real-time streaming data ingestion, delimiting, deduplication, and recalibration for highly accurate ML outcomes.
- Autoscaling for ML Workloads: Managed large-scale batch jobs, job queues, and demand surges through dynamic compute scaling.
- Hypothesis A/B Testing: Designed and executed experiments with careful consideration of variable selection, control groups, and segmentation.
