Experience

Designed and shipped an LLM-based agent that automates schema mapping, the manual step in onboarding new systems, by extracting patterns from 250+ system schema documents. Cut manual analysis time by 60% at 99%+ mapping accuracy.

Own product vision and roadmap for a real-time platform integrating 250+ systems at 500 TPS peak, prioritizing across 4 internal domains (Fraud, Risk, Servicing, ML) and aligning 3 engineering teams and 2 business units.

Defined guardrails (rate limiting, audit logging, fail-safes) with Compliance and Risk so the platform met regulatory requirements, and set the balance between new delivery and technical debt with engineering leads.

Owned strategy and launch of a self-serve Investor Portal for HNI/UHNI clients across multiple business lines and geographies. A phased go-to-market with sales, marketing and operations reached 30% adoption in month one and deflected 720 routine advisor calls a month.

Ran 20+ research sessions with portfolio managers, advisors and investors and turned the findings into journey maps that set the roadmap and MVP scope with UX, Compliance and Engineering.

Tested navigation and information architecture with users alongside UX, driving 3 major iterations of the portal's Figma prototypes.

Delivered cash-withdrawal automation, multi-account portfolio views, real-time balance checks and self-serve reporting with engineering, QA and UAT teams.

Built adoption dashboards (engagement, feature use, task success, error rates, segment penetration) and used them to find friction points and reprioritize the backlog.

Led PI planning for the investment product line, resolving competing priorities and cross-team dependencies across 3 geographies without formal authority.

Re-architected the register-search system for insider-trading compliance monitoring, cutting queries from multiple seconds to near-instant while keeping audit trails complete.

Built production systems for compliance monitoring and bond-issuance operations in SQL and Python, with APIs and event-driven data pipelines.

Worked directly with business stakeholders to diagnose pain points, recommend solutions and then build them, which grew into the move to product management.

Defined requirements with stakeholders for issuing bond tokens on blockchain.