xAI / AI Tutor — AI Training & Evaluation
Grok Voice
Helped design and refine the RLHF pipeline for Grok Voice as part of a small team of software and audio engineers, contributing to evaluation criteria, rubrics, workflows, standards, and more.
01Selected work
My Work
Professional systems, production debugging, practical automation, and embedded audio—each project shows a different part of how I think.
Front-end Technical Lead
CodeSwtch Capital 360
Contributed 250+ hours to the architectural designing, modelling, prototyping, deployment, configuration, integration, and debugging of a sponsor-backed web-application running on DigitalOcean with Docker, PostgreSQL, and MongoDB.
A practical solution to an impractical inbox
Email Organizer
A Python desktop application for organizing your inbox, featuring sender scanning, organization rules, unsubscribe workflows, and AI-assisted categorization. Tested and worked on a personal inbox with more than 27,000 emails.
Software, hardware, and sound
STM32 MIDI Player
An embedded MIDI player written in C for an STM32 Nucleo board, combining file parsing, HAL libraries, breadboarding, hardware interfaces, and digital audio playback.
02Experience
Hands-on experience with frontier AI.
Direct experience with leading AI companies, focusing on evaluation strategy, quality systems, technical review, and working closely with model developers.
01November 2024 — November 2025
xAI
AI Tutor - Voice/Audio
Remote / ContractHelped design and refine the RLHF pipeline for Grok’s voice model team of software and audio engineers. Worked with engineering leads and researchers on data-collection strategy, evaluation criteria, quality assurance, reviewer workflows, and audio-quality standards.
- Contributed to rubric design, model-response evaluation, reviewer calibration, and scalable validation systems.
- Evaluated voice output for naturalness, intelligibility, room noise, clicks, clipping, distortion, loudness consistency, and other perceptual defects.
- Identified model failure modes and helped improve evaluation consistency across large-scale review workflows.
- Contributed to Grok feature-development efforts for the X platform and standalone application while maintaining consistently strong quality scores.
02May 2024 — November 2024
Scale AI
Coding Expert for AI Training
Remote / ContractContributed to Meta AI model evaluation through Scale AI by developing challenging programming tasks, analyzing model responses, identifying inaccuracies, and providing detailed feedback.
- Applied software-engineering knowledge to supervised-learning and structured evaluation workflows.
- Tested coding and reasoning performance, documented failure modes, and recommended concrete improvements.
03Engineering foundation
My studies and technical foundation
In my five years studying software engineering at RIT, I've built a foundation that's hard to compete with.
Rochester Institute of Technology
B.S. Software Engineering
Final year of RIT’s five-year Software Engineering program.
August 2021 — December 2026 / Rochester, NYSoftware Design
Architecture, object-oriented design, patterns, subsystem boundaries, maintainability, scalability, and technical tradeoffs.
- Architecture
- Design patterns
- Requirements
- Testing
Algorithms & CS
Data structures, runtime analysis, dynamic programming, graph algorithms, search, logic, and correctness.
- Graphs
- Dynamic programming
- A* / BFS
- Runtime analysis
Web & Databases
Responsive interfaces, REST services, database-backed applications, API integration, and full-stack systems.
- React / Next.js
- Angular
- PostgreSQL
- MongoDB
Systems & Delivery
Embedded programming, production debugging, containers, remote deployment, environment configuration, and secure systems thinking.
- C / C++
- Docker
- Linux / SSH
- STM32
Languages
- Java
- C
- C#
- C++
- Python
- JavaScript
- TypeScript
- SQL
Web & data
- React
- Next.js
- Angular
- REST APIs
- PostgreSQL
- MongoDB
DevOps & tools
- Docker
- Docker Compose
- Git
- Linux
- SSH
- DigitalOcean
AI & quality
- Model evaluation
- RLHF
- Rubric design
- Voice AI
- Gen AI
04Voice AI & evaluation
Training for and evaluating naturalness in synthesized voice models
Voice models have two products at once: the behavior of the model and the quality of the audio. My background in both software and audio engineering lets me reason about both.
Input / Raw Voice SampleEvaluation PassOutput / Calibrated
NoiseClippingArtifactsProsodyIntelligibilityNaturalness
Model behavior
- Instruction following
- Naturalness and relevance
- Conversational consistency
- Comparative evaluation
- Failure-mode analysis
Audio quality
- Noise, clicks, clipping, and other sound artifacts
- Distortion and loudness
- Pronunciation and prosody
- Intelligibility
- Speaker consistency
Evaluation systems
- Rubrics and guidelines
- Reviewer calibration
- Error taxonomies
- Quality-control sampling
- Scalable review workflows
05About
My background
I’m a fifth-year software engineering student at Rochester Institute of Technology with experience across AI evaluation, software design, web development, machine learning, embedded systems, audio engineering, and more.
At xAI, I helped refine evaluation and quality workflows for Grok Voice, working alongside software and audio engineers, engineering leads, and researchers. At Scale AI, I applied programming knowledge to model evaluation work supporting Meta AI.
Outside those roles, I’ve built full-stack applications, automation tools, and embedded audio systems using Python, Java, C, C++, TypeScript, SQL, Docker, and modern web frameworks.
I’m looking for software engineering work where careful implementation, strong quality standards, and expert-level problem solving all matter.