Satyam Chouksey

I’m a Senior Full Stack Engineer with 3+ years shipping production AI safety infrastructure and SaaS platforms end-to-end. I focus on turning complex GenAI risk into reliable systems—guardrails, evaluation pipelines, and observability that teams can trust.

At Tumeryk, I’m building LLM guardrails, evaluation workflows, and real-time metrics. I engineered an AI guardrail testing agent that auto-evaluates LLM policy compliance across risk categories and outputs a model trust score, replacing manual safety review.

I also architected the base policy enforcement layer that gates every custom policy through organizational safety constraints—the core primitive downstream guardrails build on. Alongside that, I shipped an automated PII/sensitive-data validation rail to prevent confidential data exposure in live AI interactions.

Previously at Webledger, I led engineering and architected a cloud accounting platform in Adonis.js + Vue.js that cut monthly close cycles by 30% through automated reconciliation and unified cross-department data access. I enjoy owning end-to-end delivery—from system design and performance to CI/CD and reliable integrations—so products ship safely and improve continuously.