Credit Card Fraud Detection
PersonalEnd-to-end ML pipeline on 283K transactions (~0.17% fraud) — XGBoost, Platt-scaling calibration, FastAPI deployment.
PythonXGBoostScikit-learnFastAPI
AI / ML Engineer · Generative AI · MLOps
Building AI Solutions
Building AI Solutions
That Drive Real Impact
AI/ML Engineer at Capital One with 4+ years of experience building scalable, secure, and data-driven systems across fraud detection, Generative AI, and MLOps — at fintech and large-scale enterprise.
CapitalOne
AI / ML Engineer
12M+
Transactions Processed
Daily, at Capital One
< 80ms
p99 Inference Latency
Real-time Fraud Scoring
99.95%
Model Reliability SLA
Monitoring & Drift Detection
31%
Fraud Losses Reduced
Capital One Platform
Technologies I Work With
Py
PythonPT
PyTorchSp
SparkKf
Kafkaaws
AWSDk
DockerK8
KubernetesLC
LangChainXG
XGBoostPg
PostgreSQLPc
PineconeGit
GitTF
TensorFlowTf
TerraformAf
AirflowML
MLflowFeatured Projects
Enterprise Financial RAG Assistant
Capital OneRAG over 2M+ compliance documents — cutting analyst research time by 46% with grounded, cited answers.
LangChainPineconeAWS BedrockClaude
Experience Highlights
Aug 2024 – Present
AI/ML Engineer · Capital One
Real-time fraud detection on 12M+ daily transactions; productionized a RAG assistant; MLOps release cycle 3 weeks → under 48 hours.
Jul 2020 – Jun 2023
Associate SWE – AI/ML · HCLTech
Document AI pipeline (1.5M+ docs/month, review ↓62%); BERT NLP ticket routing 71% → 92%; GPU inference throughput ↑3.4x.
Education
M.S. Computer Science · NJIT
Master's degree underpinning my production AI/ML and Generative AI engineering work.