Early-to-mid career ML engineer with academic and industry experiences, deploying production-grade ML systems in the cloud that drive multimillion-dollar cost efficiencies.
rahx@utexas.edu
↗
↗
GitHub
↗
Experience
Machine Learning Engineer - Operations
- Ship production-grade multimodal LLM and computer vision pipelines on AWS with Azure DevOps CI/CD and Terraform, processing thousands of images weekly and delivering 7-figure annual savings.
- Build and scale REST APIs and MCP servers for company-wide adoption, translating ambiguous, cross-team requirements into digital products that drive multi-million-dollar operational efficiencies.
- Scale OCR and generative AI document intelligence applications with a custom evaluation harness and cost-optimized batch inference, processing millions of records while cutting inference costs by 50%.
- Manage and optimize AWS infrastructure supporting business critical ML applications and real-time analytics services across thousands of internal users.
Data Analyst - Drilling, Completions, and Well Services
- Built multi-cloud ETL pipelines for field operations data, eliminating manual entry, reducing errors/missing data, and saving ~$500k annually.
- Shipped LLM Streamlit apps for drilling and completions analysts, saving ~1 hr per well on documentation and speeding reporting turnaround.
- Published Power BI dashboards on nine-figure spend and core ops KPIs, enabling faster variance detection and timely corrective action.
Business Intelligence Associate - Operations
- Automated 10 reports and 8 interactive dashboards using Presto SQL and Apache Superset, saving 14+ hours per week for nationwide last-mile logistics and user KPI tracking.
- Delivered analytics services such as A/B testing, SQL optimization, and visualization for operations.
- Maintained an on-prem compliance pipeline producing ~150,000 legal documents monthly.
Projects
A silly little CRUD app I built to track the number of benches me and a friend sat on during our trip to Seattle. The app tracks -with user consent- the coordinate and time of the bench being logged.
Tech stack: Next.js, React, Tailwind, and MongoDB.
Pacific Northwest Hiking Poly
A low-polygonal simulation of how hiking in the Pacific Northwest feels like.
Tech stack: Three.js
Skills
Tools
- ML libraries: pytorch, scikit-learn, XGBoost, ONNX Runtiime
- DevOps: Docker, Azure DevOps, GitHub, Terraform
- Databases: Experienced in Presto SQL, Oracle SQL, Microsoft SQL Server, MongoDB, postgres, DynamoDB
Certifications
- AWS Data Engineer Associate
- AWS Cloud Practitioner
Cloud
- AWS, experienced in most common resources such as Lambda, S3, ECS, ECR, etc.
- Azure, experienced in AI services and Blob storage.