Devin D.

Key Skills

Software

Other

Top Subject Matter

Athlete performance analytics

wearable IoT data

adaptive training recommendations

Top Data Types

Text

Document

Top Task Types

Function Calling

Text Generation

Diagnosis

Question Answering

Data Collection

Freelancer Overview

Senior Software Engineer at NIKE — AI-Driven Athlete Performance Platform (LLM-powered adaptive training recommendations. Brings 12+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Other, OpenAI GPT-4, and Internal. Education includes Bachelor’s Degree, Texas Tech University (2014). AI-training focus includes data types such as Computer Code, Programming, and Medical and labeling workflows including Function Calling, Text Generation, and Computer Programming.

Labeling Experience

Senior Software Engineer at NIKE — Web3-Powered Loyalty & Personalization Engine (GPT-4 personalization)

Text Generation

Led development of a GenAI + blockchain engine that personalized an e-commerce experience using LLM-generated content. The solution combined GPT-4-driven copy generation with on-chain loyalty redemption workflows. It required coordinating AI generation outputs with blockchain interactions and user identity via wallet-based login. • Generated product descriptions and CTA copy using GPT-4 • Integrated Ethereum smart contracts (ERC-20) for tokenized loyalty points • Enabled wallet-based login via WalletConnect and MetaMask • Coordinated personalization logic across AI, backend services, and on-chain redemption.

2024 - 2025

Senior Software Engineer at NIKE — AI-Driven Athlete Performance Platform (LLM-powered adaptive training recommendations)

OtherFunction Calling

Built an AI-driven platform that integrated LLM-powered agents to generate adaptive training recommendations from biometric trends. This work involved orchestrating model context, prompts, and agent behavior to produce actionable outputs for athletes and staff. The system also supported real-time ingestion and downstream consumption of large-scale IoT data to drive the recommendations. • Used OpenAI GPT-4/LLM integration for agent-driven recommendation generation • Implemented adaptive training logic based on biometric trends • Wired agent outputs into a trainer/physician dashboard experience • Operated within a microservices + real-time data pipeline environment.

2024 - 2025

Senior Full Stack Engineer at Microsoft — AI Code Review Assistant (AI DevTools)

Designed an AI code review assistant integrated with internal GitHub repositories to assist developers with pull request reviews. The effort used ML.NET and GPT-4 fine-tuning to improve review suggestions and agent responses. The work required collecting relevant code context, managing prompts/templates, and producing programming-focused feedback to support development workflows. • Integrated LangChain.NET for agent context memory and prompt templates • Used ML.NET and ONNX-related components to power or optimize model inference workflows • Produced automated PR suggestions that improved code quality metrics • Reduced developer review time by ~35% through AI-assisted feedback.

2020 - 2023

Lead Full Stack Developer at AmarilloIT — CityGov CRM with GenAI Ticket Triage

TextQuestion Answering

Built a GenAI-enabled CRM feature for automatic ticket categorization and citizen response generation using an LLM model. The solution transformed incoming text tickets into categorized outputs and generated responses for citizens, supporting operational workflows. It required integrating AI generation into backend services and coordinating data flows with a structured CRM UI. • Generated automatic ticket categories using an LLM model • Produced citizen response text via LLM generation • Integrated AI features into Express/Django backend services • Connected AI outputs to an Angular frontend with NGXS state management.

2014 - 2019

Lead Full Stack Developer at AmarilloIT — Smart Hospital AI Assistant Dashboard

OtherDiagnosis

Developed a full-stack smart hospital AI assistant dashboard that used AI agents trained on hospital-specific data. The assistant provided doctors with diagnosis suggestions based on clinical NLP inputs and agent reasoning. This required integrating trained transformer models and building interfaces for real-time clinical interaction. • Integrated Hugging Face Transformers for clinical NLP tasks • Used AI agents trained on hospital-specific data for diagnosis assistance • Delivered real-time dashboards for clinician consumption • Implemented web and mobile UI components for clinical workflows.

2014 - 2019

Show more

Education

Work History