Zoox Intelligence(Agents, MCPs, Evals, Guardrails, Connectors(RAG), Fine Tuning, Agentic SDKs and more)

AI to enhance employee experience, focusing on LLM fine-tuning, information retrieval, and constructing RAG frameworks.

• Built deep learning models (CNN, LSTM, Seq2Seq, BERT) for protein toxicity prediction and NLP research.

• Developed the RDASH platform using NLP and RAG to map 60K+ funding calls to 5K+ scholars, improving alignment and efficiency.

• Designed recommendation and compliance models for tailored funding insights and automated discovery.

• Accelerated data processing pipelines using NVIDIA RAPIDS, achieving up to 14× faster performance on GPU.

• Optimized CUDA based metrics and imputation algorithms for large-scale NLP tasks.

• Built and benchmarked XGBoost models on GPU, validating performance gains across workloads.

• Built hybrid recommendation and propensity engines to improve retention and sales outcomes in e-commerce

• Designed clustering models to identify high valued credit card holders, defaults, created fraud alerting mechanisms across cards, payments, accounts.