Sashank Kadiveti

I’m a Computer Science student focused on turning AI into reliable, real-world features—especially where compliance, fairness, or utility matter. I build end-to-end systems that combine ML/LLM capabilities with production-minded APIs, data pipelines, and UI.

In standout projects, I built an AI Lending Bias Auditor that trains an XGBoost model to replicate lending decisions and audits for disparities using Fairlearn, generating regulator-ready compliance reports. I also created ContextPack, a CLI that generates token-efficient LLM context using multi-signal ranking and RAG-style deep dives with history compression to stay within token limits.

I’m equally comfortable shipping full-stack apps and mobile experiences, from real-time civic infrastructure detection to voice-to-code Python generation deployed with containerized backends. As a CS110 teaching assistant, I reinforce core programming concepts through lab guidance and debugging—so my work reflects both technical rigor and clear, user-focused problem solving.