Eshaan Saha

I’m an engineering-focused early-career builder who turns complex enterprise inputs into standardized, governed systems. At Ernst & Young (EY), I developed Catalog Curator for software catalog normalization, automated enrichment validation using fuzzy matching and lifecycle tracking, and created an Agent Observability Platform to monitor AI agent execution and telemetry at enterprise scale.

I also built an LLM Cost & Usage Management System to track model invocations, analyze token consumption, optimize operational costs, and centralize visibility into enterprise AI workloads. Alongside this, I enjoy designing systems that improve data quality, reliability, and operational governance rather than shipping features in isolation.

Earlier, I worked as a Research Intern at IIT Jodhpur, building end-to-end EEG data pipelines in MATLAB and Python and running behavioral analysis for stop-change tasks. My projects reflect the same mindset—e.g., AssetMind uses CVE/CVSS vulnerability intelligence for risk scoring and compliance mapping, and my EEG-based attention work trained ensemble models on 14,980 samples to achieve 96% accuracy.