✅ Word Document: Click "Word" to download an editable Microsoft Word (.docx) file.
✅ PDF Download: Click "Download PDF" to instantly download a pixel-perfect A4 PDF file.
✅ Browser Print: Click "Print Dialog" if you want to print to a physical printer (remember to enable "Background graphics").
Umakant Yadav
Principal Data Scientist & Machine Learning Lead
uky171991@gmail.com
9453619260
San Jose, CA
github.com/samanthalin-ml
ABOUT
Principal Machine Learning Scientist with 8+ years developing predictive recommendation engines, LLM fine-tuning pipelines, and high-frequency anomaly detection models for Big Tech. Holds 4 AI patents in deep reinforcement learning.
EXPERIENCE
Principal ML Engineer & Team Lead
2021 — Present
Apex Neural Technologies
Designed transformer-based personalization model increasing click-through-rate by 28% and driving $42M incremental GMV.
Architected distributed inference infrastructure handling 350M daily API predictions with sub-15ms p99 latency.
Authored continuous automated model drift retraining pipelines using MLflow and Kubeflow.
Senior Machine Learning Scientist
2017 — 2021
Veritas Analytics Lab
Built fraud detection classifier scoring 12M credit card transactions daily with 99.8% precision.
Mentored 8 junior data scientists and published 5 peer-reviewed papers at NeurIPS and ICML.
EDUCATION
Ph.D. in Computer Science & Artificial Intelligence2017
Stanford University
B.S. in Applied Mathematics & Statistics2013
Carnegie Mellon University
EXPERTISE
PyTorchDeep Learning & TransformersDistributed Training (Ray/vLLM)MLOps & MLflowPythonC++Docker/KubernetesVector DBs (Pinecone/Milvus)