Open to opportunities

austin lee

@austinlee1

Message

Senior Data Scientist focused on LLMs, RAG, and scalable AI solutions.

United States

Message

What I'm looking for

I'm seeking a senior role to lead production-grade LLM/RAG and cloud-native ML initiatives in fintech or e-commerce—driving measurable business impact, mentoring teams, and ensuring compliant, scalable deployments.

I am a Senior Data Scientist with 12+ years building machine learning and AI solutions across fintech, e-commerce, and financial risk analytics.

I design and deploy LLM-driven systems and RAG pipelines that have improved credit decision turnaround by 45% and lifted approval rates while protecting portfolio health, unlocking multimillion-dollar business impact.

I architect cloud-native AI infrastructure using AWS Bedrock, SageMaker, Lambda, S3, DynamoDB, ElastiCache, alongside Pinecone and LangChain, achieving 99.9% availability and meaningful latency reductions.

I partner with product, risk, and executive stakeholders to productionize models with explainability and regulatory compliance, mentor junior scientists, and embed CI/CD and experiment tracking to accelerate safe, measurable outcomes.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

Cognizant

Apr 2021 - Present (4 years 5 months)

Spearheaded RAG pipelines integrating OpenAI, Cohere, and Anthropic to optimize SME loan underwriting, improving credit decision turnaround time by 45% and increasing approval rates by 12%; built cloud-native AI infrastructure on AWS achieving 99.9% availability and 23% lower latency.

Cohere Anthropic Pinecone Python

Data Scientist

Amazon

Jun 2019 - May 2021 (1 year 11 months)

Led customer segmentation and lifetime value modeling and deployed predictive churn models using SageMaker, improving retention KPIs by 18% and increasing cross-sell conversions by 22% through targeted campaigns.

Customer Segmentation LTV Modeling AWS Sagemaker Python SQL Redshift scikit learn XGBoost

AI Engineer

Kabbage

Oct 2014 - Jun 2019 (4 years 8 months)

Developed real-time fraud detection and deep-learning borrower risk scoring models, reducing fraud loss rate by 28% and improving default-prediction AUC from 0.78 to 0.87 while automating lending workflows to cut decision time to under 10 minutes.

Fraud Detection Kafka Spark TensorFlow Keras Python Risk Modeling

Machine Learning Engineer (Intern)

OnDeck

Jun 2013 - Oct 2014 (1 year 4 months)

Conducted exploratory data analysis on borrower behavior and prototyped logistic regression and decision-tree credit scoring models that surfaced early repayment-risk indicators; supported A/B tests that increased completed applications by 9%.