At DECCAN AI, I validated multi-turn language models and speech-to-speech criteria for Project Payphone. I benchmarked conversational logic constraints and examined contextual behavioral edge cases.
At MERCOR, I verified algorithmic logic and mathematical correctness for Project Odin. For Project Folio, I annotated multilingual documents using SuperAnnotate and tracked billable work in Insightful.
Across CrowdGen, Handshake AI, OneForma, and Outlier AI, I evaluated text compliance, multimodal reasoning, search relevance, model safety, and RLHF alignment. At Outlier AI, I reached Tier 2 based on data audit metrics and linguistic task reviews.
My aerospace work includes testing a contra-rotating ducted propeller engine and, as an intern at Government Tool Room & Training Centre (GTTC), designing scale models and running structural and thermal fatigue stress meshes. I also co-designed and fabricated a pneumatic landing gear retraction assembly.
