At OneForma, I evaluated LLM and multimodal outputs across text, image, audio, and bounding-box tasks. I annotated more than 3,000 tasks with zero critical guideline violations and detected hallucinations and reasoning errors across six or more task types.
My evaluation work included assessing 50+ LLM and benchmark output pairs each month for accuracy, coherence, and task relevance. I contributed to dataset quality improvements used for benchmark evaluation.
At PT PP (Persero) Tbk, I supported structural monitoring on an IDR 2T+ dam project and delivered daily and weekly progress and deviation reports. I also assisted with structural calculations and design validation.
Alongside my engineering work, I’ve produced geopolitical analysis for an audience of 5,000+ and created social media content. I’m completing the IBM Generative AI Engineering Professional Certificate, including its Machine Learning with Python module.

