Ahmad Dzulfikar
Hi there, I’m Ijul—
Information Systems student at Telkom University and AWS Certified Data Engineer. I test whether AI models follow instructions and reason well, and I write up where they fail so they can be improved.
AI evaluation now, esports casting before that
Abaka AI
Evaluated how well AI systems understood and followed instructions on multimodal tasks such as image editing, text generation and research synthesis. I tested responses with ambiguous prompts and edge cases, and wrote structured feedback on where an output missed what the user meant.
Invisible Technologies
Evaluated AI outputs on varied tasks for accuracy, reasoning, interpretability and instruction-following, and judged how closely each one matched what the user expected.
Dunia Games
Live commentary and strategic analysis for League of Legends and Teamfight Tactics community tournaments. I explained drafts, team compositions and strategy for newcomers and longtime fans alike.
Diffarenz
Live commentary and tactical analysis for a national League of Legends tournament, explaining complex plays so the audience could follow them.
Backed by work experience and an AWS certification
An associate-level AWS certification for building and running data pipelines.
For over a year at Abaka AI and Invisible Technologies, I checked how well multimodal AI follows instructions and reasons, and found where it breaks.
I use SQL, Python and AWS to ingest, store and process data, and I hold the AWS Data Engineer certification.
I’ve explained complex gameplay live on a national esports broadcast and hosted a university cultural event as MC. I bring the same clarity to the feedback I write on model behaviour.
Telkom University · ’27
I’m a final-year Information Systems student at Telkom University with hands-on experience in AI evaluation and language-model assessment. I like finding the exact point where a model misreads what someone asked, then explaining it clearly enough that it can be fixed.
Off the clock I’ve cast League of Legends and TFT tournaments, and served as MC and event committee member for Nippon Bunka-Bu, the university’s Japanese culture club. Explaining a draft pick to a live audience turned out to be good practice for explaining a model failure.
Open to roles and projects in AI evaluation, data engineering and information systems.