Sep 2019 – Present · 6 yrs 7 mos

• ► Leading the buildout of an AI-first recovery services business, blending proprietary conversational AI, workflow automation, and human ops to deliver scalable, outcome-driven resolutions.
• ► Led a team of ~10 product managers and designers to build and scale an Agent-Application platform, accelerating U.S. market entry through faster customization velocity (scaled up to 100+ customers), contributing to $XM ARR in the initial ~18 months.

Jun 2018 – Sep 2019 · 1 yr 3 mos

• ► Hired and built various teams (Solutions, Product, and Conversational User Experience teams). Played the role of the first Solutions Product Manager and Conversation Designer/Writer in the initial days.
• Hired/onboarded the initial group of Solutions Product Managers
• Established the CUX team by hiring the early team of conversation designers and writers
• Created and taught a course (CD-102) on "Voice-centric Conversation Design" to onboard the CUX team
• ► Launched a multilingual [10+ languages] crowdsourcing platform [Amazon Mechanical Turk equivalent for India], accelerating the development of our voice roadmap and reducing the projected cost by 300-400%.
• Reduced CAC by 80% using influencer marketing [YouTube/Instagram partner affiliation] before it went mainstream in the country
• Managed customer support and product operations (payments)
• ► Played the product manager role in the small skunkworks team leading the execution during the organization’s pivot from chat to voice.
• Coordinated the technical and growth initiatives to help build the roadmap for Vernacular.ai’s voice conversational experience

Jun 2017 – May 2018 · 11 mos

• Led the design and implementation of the Event-Driven architecture (EDA) for the voice-services project.
• Introduced Kubernetes in the internal infrastructure
• Finite State Machine based workflow orchestration
• Integrated with cloud telephony providers
• Golang/Python/Kubernetes
• Implemented voice data collection tools to be used further in different services.
• Application stack to crowdsource and auto tag audio data
• Streaming based voice recognition middleware service used by Web SDK that integrates with ASR and the storage layer
• Designed (architecture) and developed an internal tracing and analytics system across the infrastructure. It involved implementing a custom logging format (and loggers) compatible with existing services as per business requirements, and batch processing based data-collectors to power the analytics, using Python/NodeJS/ElasticSearch.
• Developed the ChatBot Web SDK project from scratch using React/Redux/ES6. It's an embeddable widget supporting multiple messaging features.