[AI Platform & Infrastructure] -Unified AI Platform: Spearheaded a centralized API and billing layer to securely scale LLM integrations (OpenAI, Azure, AWS Bedrock) across all internal product teams. -Accelerated Delivery: Eliminated redundant engineering efforts, increasing feature delivery velocity and seamlessly integrating newly acquired products onto the core platform. -Standardized RAG Architecture: Launched an enterprise-grade Knowledge Service with 'Bring Your Own Knowledge' (BYOK) capabilities, consolidating disparate internal RAG pipelines into a single, secure solution. [Commercialization & Strategy] -AI Monetization Strategy: Designed and implemented a token-to-credit usage-based pricing model, optimizing unit economics and heavily reducing customer adoption friction. -Enterprise Cost Control: Shipped the AI Credit Report console, equipping enterprise administrators with prompt-level telemetry to confidently forecast costs and manage their AI investments. -Agent Adoption & Growth: Directed the product roadmap for Knowledge and Document Discovery Agents, driving robust market adoption and sustained month-over-month growth. [AI-Native UX & Trust] -Enterprise AI Design System: Owned the Coupa Navi UX layer, standardizing streaming responses, transparency cards, and guardrail messaging across the entire corporate AI portfolio. -Trust & Safety Optimization: Embedded deep explainability and safety guardrails directly into the user experience, directly resulting in elevated user trust and CSAT scores. -Customer-Centric Road mapping: Conducted strategic discovery interviews with enterprise clients, translating qualitative feedback into actionable strategies for transparency and monetization.

Designed and shipped a Test Management MCP server connecting 16+ internal tools to LLM agents as part of Engineering platform LLM Performance Optimization: Diagnosed and resolved critical latency bottlenecks, significantly improving overall agent response times and system scalability. Automated Quality Infrastructure: Built a multi-tier performance testing suite (C#/.NET, Vue.js, AWS) with real-time analytics, replacing ad-hoc ML quality checks with robust, automated monitoring.

ML Testing Framework : Designed "Monarch," a comprehensive ML testing framework (regression, API, load) adopted cross-functionally as the primary internal quality gate for AI releases. Full-Stack Telemetry: Shipped a microservices-based test tracking platform (C#/.NET, Vue.js, AWS) providing real-time reporting and visibility across all engineering and QA teams. AI Capacity & SLA Planning: Established foundational latency benchmarks for enterprise AI systems, informing strict SLA definitions and mitigating architectural bottlenecks prior to production deployment.