DeVon Rasch, M.S.

(512) 516-1383 ยท devonrasch@gmail.com

Innovative Data Scientist and AI Engineer with 8+ years of experience specializing in scalable infrastructure and machine learning. Proven track record of deploying production-grade AI agents and optimizing large-scale data pipelines. Expert in leveraging LLMs, vector databases, and image processing to drive multi-million dollar cost savings and operational efficiency.

Proficient in modern AI development workflows including Claude Code development, Claude Code MCPs, and Cursor IDE for rapid engineering. Experienced in fine-tuning Hugging Face open-source models, utilizing Ollama for local model optimization, and Databricks MLflow for managing the machine learning lifecycle and Snowflake's Cortex.

Experience

Core AI/Data Engineer

SciPlay
  • Architected and deployed autonomous AI agents into production to automate complex decision-making workflows, enhancing operational responsiveness.
  • Generated over $7,000,000 in storage cost savings through advanced data lifecycle management and compression algorithms.
  • Managed the configuration and fine-tuning of Large Language Models (LLMs) to improve model accuracy for domain-specific tasks.
  • Developed high-performance retrieval systems utilizing Vector Databases to enable semantic search capabilities across massive datasets.
January 2023 - Present

Machine Learning Engineer - AI Ops (Contract)

Apple
  • Scaled anomaly detection system from 10 to 200 pipelines while reducing SLA times from 20 to 1 minute.
  • Developed a self-serving API enabling automated pipeline onboarding for customized anomaly detection.
  • Built a data classification system for unsupervised machine learning models, preventing DDOS attacks on Apple TV and fraudulent Apple Pay transactions.
  • Utilized AI-driven image processing techniques (OpenCV) to automate visual data verification, reducing manual oversight by 40%.
April 2022 - September 2024

Data Engineer

Saatva
  • Engineered ETL pipelines integrating multiple advertising platforms (Facebook, Twitter, Snapchat) into AWS Athena.
  • Developed automated data ingestion solutions using Python and PySpark for marketing analytics.
  • Implemented hourly data refresh cycles using AWS Glue scripts for real-time marketing insights.
June 2021 - July 2022

Full Stack Engineer (Contract)

AT&T
  • Led development of Northstar, a full-stack application consolidating multiple legacy data sources into a single source of truth.
  • Architected a data validation system to identify mismatched types and synchronization issues.
  • Designed and implemented ETL processes based on user-defined forms and requirements.
January 2021 - April 2022

Data Scientist

Degree Analytics
  • Analyzed student success patterns using IoT sensor data and demographic information to identify key success factors.
  • Managed AWS infrastructure, including S3, SFTP, and IAM controls for university data onboarding.
  • Created data visualizations using AWS QuickSight to communicate insights to university stakeholders.
October 2019 - December 2020

Associate Technologist

Merck
  • Developed MVP projects, including an innovative LLM-powered (Rasa) chatbot for SharePoint integration.
  • Led international RPA initiatives to automate business processes.
  • Collaborated with cross-functional teams to deliver rapid prototypes for innovation projects.
January 2019 - October 2019

Robotic Process Automation Developer (Contract)

Merck
  • Automated SAP discount verification process for 1,000 vaccination purchases, resulting in $250,000+ savings within the first 6 weeks.
  • Implemented OCR machine learning automation solutions for international teams.
October 2018 - January 2019

Data Scientist (Intern)

Degree Analytics
  • Created a facility usage heat map feature using PySpark and Seaborn for visualizing campus activity.
  • Conducted machine learning research to address university-specific feature requests.
May 2018 - August 2018

Education

Eastern University

Master of Science, Data Science
2025

Huston-Tillotson University

Bachelor of Science, Mathematics
Austin, Texas
2019

The University of Texas, Austin

Data Engineering / Data Scientist Certificate

Austin Community College

Associate of Science, Mathematics

Skills

Cloud Development
  • AWS EC2 — Virtual compute instances
  • AWS EKS — Managed Kubernetes clusters
  • AWS ECR — Container registry
  • AWS ECS — Container orchestration
  • AWS S3 — Object storage & data lakes
  • AWS Lambda — Serverless compute
  • AWS Athena — Serverless SQL querying
  • AWS Glue — ETL & data cataloging
  • AWS Kinesis — Real-time data streaming
  • AWS Redshift — Data warehousing
  • AWS SQS — Message queuing
  • AWS IAM — Identity & access management
  • AWS QuickSight — BI & dashboarding
  • AWS CloudWatch — Monitoring & logging
  • Azure Container Registry
  • Terraform — Infrastructure as code
Data Science
  • Pandas & NumPy — Data manipulation
  • SciPy — Scientific computing
  • Seaborn & Matplotlib — Data visualization
  • Polars — High-performance DataFrames
  • Apache Spark & PySpark — Large-scale processing
  • Feature engineering & data cleaning
  • Statistical analysis & hypothesis testing
  • Time series analysis & forecasting
  • Anomaly detection systems
  • Delta Live Tables — Streaming pipelines
  • Data lake & data warehouse architecture
  • A/B testing & experimentation
Artificial Intelligence
  • LLM configuration & fine-tuning (GPT, Claude, Llama, Mistral)
  • Hugging Face — Open-source model fine-tuning
  • Ollama — Local model optimization
  • OpenAI API & Anthropic Claude API
  • Vector databases (Pinecone, ChromaDB, FAISS)
  • Text embeddings & sentence chunking
  • RAG — Retrieval-Augmented Generation
  • AI agent orchestration (LangChain, AutoGPT)
  • Multi-agent systems & agentic workflows
  • Human-in-the-loop AI processes
  • Prompt engineering & chain-of-thought
  • Snowflake Cortex — AI on structured data
Machine Learning
  • TensorFlow & PyTorch — Deep learning frameworks
  • Scikit-learn — Classical ML algorithms
  • XGBoost & LightGBM — Gradient boosting
  • Ensemble methods & model stacking
  • CNNs, RNNs & Transformer architectures
  • Image processing — OpenCV & YOLOv3
  • Natural Language Processing (NLP)
  • Reinforcement learning & Q-learning
  • Classification, regression & clustering
  • Unsupervised learning & dimensionality reduction
  • Model evaluation, validation & hyperparameter tuning
  • Computer vision & object detection
Programming Languages
  • Python (Advanced)
  • SQL (Advanced)
  • PySpark
  • JavaScript & TypeScript
  • HTML & CSS
  • R
Software & Tools
  • AWS SageMaker — ML model training & deployment
  • Databricks — Unified analytics platform
  • MLflow — ML lifecycle management
  • Snowflake — Cloud data warehousing
  • Apache Airflow — Workflow orchestration
  • Docker — Containerization
  • Kubernetes — Container orchestration
  • Tableau — Business intelligence & visualization
  • Jenkins — CI/CD automation
  • REST APIs & FastAPI
  • PostgreSQL & Supabase
  • Jupyter Notebooks
  • Git & GitHub
  • UiPath — Robotic process automation
Procedures & Methodologies
  • CI/CD pipeline development & deployment
  • RAG system design & implementation
  • LLM fine-tuning workflows
  • AI agent development & orchestration
  • Human-in-the-loop AI process design
  • Robotic Process Automation (RPA) & OCR
  • ETL pipeline design & data modeling
  • Data lake architecture & lifecycle management
  • Agile & Scrum methodologies
  • SFTP & third-party data integration

Interests

Apart from being a data scientist and AI engineer, I enjoy most of my time being outdoors. In the summer, I am an avid skateboarder. You can find me in Durango, Colorado longboarding down mountains at high speeds avoiding the Texas heat. I love to explore when I am not with family and love brand new hip hop.

When indoors, I follow a number of data science blogs and enjoy building things. I have created my own reinforcement learning trading robot and continue to push it further. I'm fascinated by how AI can learn to understand markets at all levels and hope to one day let it run fully autonomously.

Awards & Certifications

  • Databricks Certified Data Engineer Associate — Credential ID 70165650
  • Level 3 International RPA Developer UiPath Certificate
  • Andrew Ng's Stanford Machine Learning Course - Coursera Certification
  • 1st Place - Huston-Tillotson University - Emerging Tech Competition 2017
  • 2nd Place - Huston-Tillotson University - Diversity Hackathon 2017
  • 2nd Place - HBCU Battle of the Brains Hackathon - Home Depot Tech Competition 2018
  • 1st Place - Mentor - Diversity Hackathon 2018
  • 3rd Place - HomeAway - Hackathon 2018

Projects

Pocket LLM

A UI that lets you insert PDFs or text files into a vector database and use an LLM to draw context from the vector embeddings of those files.

View on GitHub

Multi-GPU Setup

An environment to allow TensorFlow or PyTorch to leverage multiple GPU devices on a single machine and benchmark usage across GPUs when training a new network.

View on GitHub

LangChain Fundamentals

An intro to LangChain fundamentals covering core concepts for building LLM-powered applications.

View on GitHub

NLP Mastery

A collection of NLP techniques and experiments covering the breadth of natural language processing methods.

View on GitHub