Open to opportunities

Samuel Salem

@samuelsalem2

I’m a Data Architect specializing in enterprise data platforms and production LLM/ML systems with end-to-end RAG and MLOps expertise.

United States

Message

What I'm looking for

I’m looking to architect and ship production-grade data and AI systems—especially LLM/RAG platforms—where I can own end-to-end pipelines, MLOps (CI/CD, monitoring, drift detection), and build reliable, compliant solutions that measurably improve outcomes.

I’m a Data Architect with 8+ years designing enterprise data platforms and building/deploying ML/DL models across NLP, Predictive Analytics, and Generative AI. I bring deep expertise across data architecture, data science, and LLM engineering—especially RAG-based retrieval systems (embeddings, semantic search, reranking) and Transformer architectures—with end-to-end ownership from pipeline design to production deployment.

In healthcare-focused environments, I architected LLM-powered, agentic AI systems for clinical decision support, care coordination, and anomaly detection across 12 production settings, supporting 150+ clinicians. I’ve delivered enterprise data models, ETL/ELT pipelines, and RAG architectures across 2M+ clinical and claims records—improving accuracy by 28% and cutting hallucinations by 45%—while reducing clinician triage/documentation time by 30% across 25+ workflows; I also directed HIPAA-compliant infrastructure and APIs (FastAPI, Flask, SQL/NoSQL) supporting 500K+ requests/day. Previously, as a Senior AI/ML Engineer and AI/Data Engineer, I led end-to-end transformer/LLM and NLP initiatives across 1.5M+ patient records, improving predictive accuracy by 22%, cutting model deployment time from weeks to under 48 hours, and reducing model latency by 50% to support 10K+ predictions/hour.

Experience

Work history, roles, and key accomplishments

Current

Data Architect

Current

Midwestern

Oct 2024 - Present (1 year 9 months)

Architected scalable data platforms and pipeline infrastructure for LLM-powered, agentic AI systems supporting clinical decision support, care coordination, and anomaly detection across multiple production healthcare environments. Designed enterprise data models, ETL/ELT pipelines, and RAG architectures and established HIPAA-compliant data infrastructure and governance.

Data Architecture LLM Agentic AI RAG Embeddings Semantic Search Reranking Data Modeling ETL Data Governance ML Evaluation fastAPI Flask SQL NoSQL HIPAA compliance Anomaly Detection API Development

Senior AI/ML Engineer

Pintech

Mar 2022 - Sep 2024 (2 years 6 months)

Designed and deployed machine learning and deep learning models (NLP, classification, forecasting) for healthcare applications. Built end-to-end ML pipelines and production ML APIs/microservices, and implemented MLOps practices including CI/CD, model monitoring, and automated retraining.

Machine Learning Deep Learning NLP Forecasting Python TensorFlow PyTorch scikit learn Feature Engineering Hyperparameter Tuning Model Validation fastAPI Flask Docker Kubernetes Microservices CI CD Model Versioning Model Monitoring

Senior AI/Data Engineer

Peer AI

Jan 2020 - Feb 2022 (2 years 1 month)

Led a team designing, training, and deploying deep learning and NLP models for clinical risk prediction, patient outcome forecasting, and diagnostic support. Built transformer/LLM models on clinical text and EHR data and implemented MLOps for CI/CD, model versioning, retraining, and monitoring, including real-time inference services and model validation for clinical safety and regulatory readiness

Transformer Models LLM Fine Tuning Deep Learning NLP MLOps CI CD Model Versioning Model Monitoring Model Validation

ML Data Scientist

Combinaut

Jun 2018 - Dec 2019 (1 year 6 months)

Applied statistical and machine learning methods to healthcare data (structured EHR/claims and unstructured clinical notes) to identify patient risk factors and outcome trends. Evaluated model performance with standard classification/regression metrics and designed A/B tests and cohort studies to measure intervention impact.