> Initializing portfolio system...

> Loading modules...

> System ready.

Henish Patel
> whoami

Henish Patel

CS + Data Science @ University of Utah (May 2027)

I build backend tools, data workflows, and applied AI systems.

Current work: self-hosted LLM infrastructure, RAG workflows, FastAPI services, and model evaluation.

01. / About

Background

I am an engineering student pursuing dual degrees in Computer Science and Data Science at the University of Utah.

My technical foundation spans full-stack development, distributed data workflows, and applied machine learning. I am heavily focused on designing resilient backend architectures and deploying intelligent systems—whether that involves architecting secure, self-hosted LLM inferences, automating complex ETL pipelines, or building robust RAG architectures for internal data retrieval.

I am actively seeking Software Engineering, AI Engineering, and Data Science internship opportunities for Summer 2027 where I can contribute to mission-critical infrastructure and continue solving challenging technical problems.

02. / Experience

Work History

AI Systems Intern

May 2026 – Present

Management & Training Corporation

  • Architecting a self-hosted AI inference stack with Ollama and Docker, deploying open-source LLMs for secure on-premise reasoning with zero external API exposure.
  • Building retrieval-augmented generation (RAG) pipelines with vector embeddings to ground self-hosted models in internal documents, improving answer relevance with full on-premise data privacy.
  • Designing evaluation harnesses benchmarking local model latency, accuracy, and resource utilization to accelerate retrieval and prompt iteration.

Full Stack Software Engineer

Oct 2025 – Present

mySpecSheet

  • Architected an AI-integrated dashboard utilizing VSCode OSS and TypeScript; implemented custom LSP extensions to reduce developer context switching by 38%.
  • Engineered a “vehicle-as-a-repo” architecture utilizing Merkle-tree versioning and MCP Servers for secure, immutable repair history and sandboxed data manipulation.
  • Developed a WebSocket orchestration layer for real-time telemetry synchronization with sub-50ms latency; implemented OAuth2/RBAC ready for 1,000+ concurrent service nodes.

SUDO Software Platform Services Intern

July 2024 – April 2026

University of Utah

  • Optimized GIS geospatial indexing using React, Node.js, and R-trees, improving spatial data query performance by 27% and reducing client-side rendering latency by 140ms.
  • Architected ServiceNow ETL pipelines to automate performance auditing, reducing processing time from 4 hours to 15 minutes and ensuring 100% data consistency across relational schemas.

Data Analyst Intern

May 2024 – August 2024

University of Utah Health Facilities Management

  • Led zero-downtime migration of legacy reporting workflows to AWS (S3, EC2) with MD5 checksum validation across 500GB+ datasets, ensuring total data integrity.
  • Engineered modular ETL pipelines in Python (Pandas) automating reconciliation of 15,000+ monthly records.

Information Technology Support Staff

May 2023 – July 2024

Weber State University

  • Provided technical IT support and maintained campus computing environments, troubleshooting hardware and networking protocols.
03. / Engineering

Selected Projects

NVIDIA Grace-Blackwell DGX Spark Cluster

Built a DGX Spark Cluster utilizing ConnectX-7 200Gb/s interfaces and RoCE v2 (RDMA) for peer-to-peer GPU memory access, creating a unified 256GB+ VRAM pool. Utilized CUDA accelerated libraries to offload workloads, achieving 50x speedups.

PythonCUDADockerSpark

Stripe Agentic Billing Extension

Used the Stripe V2 EventStream API to enable real-time, token-based monetization for autonomous agents. Implemented a Redis ZSET-backed event aggregator with Lua scripting to atomize credit burndown, reducing API overhead by 85%.

Node.jsRedisPostgreSQLStripe API

Formula 1 Race Prediction

Engineered a machine learning model predicting F1 race outcomes by analyzing historical telemetry, driver statistics, and weather data.

PythonScikit-learnPandas

EPL Fixture Results Prediction

Built a predictive model forecasting English Premier League match results using advanced feature engineering and machine learning algorithms.

Machine LearningPythonData Viz

Image Editor & JSON Connectivity

Developed a full-stack application enabling comprehensive image manipulation with robust JSON-based state saving and network connectivity.

ReactNode.jsJSON
04. / Tooling

Technical Skills

Languages

PythonJavaC++C#CJavaScriptSQLBashAssembly

AI / ML / Data

Machine LearningRAGLLMsPandasNumPyScikit-learnTensorFlowPyTorchJupyterData Visualization

Backend & Infrastructure

FastAPIDockerPostgreSQLSQLiteAWS EC2AWS S3CI/CD

Frontend & Tools

ReactTypeScriptTailwind CSSGitGitHubVS CodeTableauPower BI
05. / Credentials

Certifications

  • Google IT Automation with Python
  • Google AI Agents Course
  • Oracle Database SQL
  • IBM Data Science Professional
  • Generative AI Fundamentals
  • NetQ Deployment and Installation

Curriculum Vitae

Download my complete resume to view my academic background, full project details, and comprehensive technical experience.

View Resume PDF