Aarush
Agarwal

Researching dynamic Mixture-of-Experts architectures under Chenyan Xiong, developing adaptive strategies that expand model capacity on out-of-distribution data while mitigating reasoning degradation in continual pretraining.

Designing autonomous research agents that propose, execute, and evaluate model-adaptation experiments across tasks and modalities including vision, clinical, and financial time-series data, iteratively committing variants that improve reasoning and task-benchmark performance over dense backbones.

Search Relevance: Building ranking systems for Shopify's commerce-search stack, including fine-tuning and distilling Liquid AI LFM2.5 re-rankers for low-latency inference on custom Triton serving infrastructure. Reduced p99 inference latency by 5% (20 ms → 19 ms).

Merchant-Aware Ranking: Designed and implemented an auxiliary merchant-aware training objective for merchant-intent queries. Improved a core merchant-search relevance metric by 10% by helping the ranker surface a merchant's first-party catalog ahead of third-party resellers.

Win/Loss Analysis: Identified data and supervision mismatches by comparing production re-ranker wins and losses against LLM reference judgments, establishing the need for new training data.

Query Rewriting: Developing a unified Storefront/Shop/Catalog query-rewriting model (+0.5 NDCG) and improving non-English rewrites.

Co-organized Felicis × CMU VentureHacks, selecting 53 builders from 150+ applicants for an eight-hour build sprint that produced 20 project demos; awarded $10K+ in prizes and hosted an AMA with a Skild AI founding researcher.

AI Venture Research: Conducted startup diligence and market research on emerging AI companies and technical market shifts alongside the Felicis team.