AI Trainer — AfterQuery & Revelo · 2026
AI Training & Evaluation
AI training work spanning data annotation, model evaluation, and testing AI-generated code, drawing on my roles at AfterQuery and Revelo.
Problem
AI-generated responses and code need careful evaluation to identify errors, edge cases, and gaps in quality.
Contribution
At AfterQuery, I work on data annotation and analyze model performance metrics. At Revelo, I evaluated, tested, and debugged AI-generated code and contributed RLHF-oriented feedback.
Outcome
Contributed annotated data, code evaluations, and feedback to AI training workflows across both roles.
Technical decisions
- Check generated code through testing and debugging to ground feedback in observable behavior.
- Use structured feedback and consistent evaluation criteria to make quality issues clear.
- Document edge cases that reveal gaps in model behavior.
AfterQuery — AI Trainer
Since May 2026, my work has included data annotation, reviewing model performance metrics, surfacing edge cases, and contributing feedback on evaluation criteria.
Revelo — AI Trainer
From January to May 2026, I evaluated, tested, and debugged AI-generated code and supported training workflows through RLHF-oriented feedback.
Scope & limitations
This is an overview of professional responsibilities across two roles, rather than a standalone application. No quantitative model-improvement claim is made.