Apr 2023 – Present · 3 yrs 2 mos

• At Observe.AI, I’ve been driving the architecture and development of multiple real-time, enterprise-grade AI products, helping redefine how contact centers leverage AI in live customer interactions. My work blends deep technical architecture, team leadership, and product vision to deliver scalable systems that operate reliably at massive scale.
• Real-Time Agent Assist -
• Spearheaded the end-to-end development of the Real-Time Agent Assist product — from idea inception and architecture to production launch.
• Built and led a 12-member engineering team, including backend, frontend, and an engineering manager, to own and scale the product.
• Designed and implemented a cost-efficient, horizontally scalable architecture supporting thousands of concurrent sessions.
• Engineered a novel audio acquisition pipeline that captures both microphone and speaker audio from user endpoints for live transcription.
• Developed a streaming phrase-matching algorithm for real-time detection of conversational cues and compliance moments.
• Implemented dynamic prompt generation using RAG (Retrieval-Augmented Generation) for context-aware agent guidance.
• Rust-based service that reads audio streams from Linux proc files and streams them to backend processing systems
• Achieved end-to-end latency under 1 second, delivering real-time insights during live calls.
• Architected a high-performance transcription delivery service capable of 15,000+ concurrent real-time sessions with P100 latency of just 138ms on single CPU core.
• Built Major CCaaS integrations
• Supervisor Assist -
• Designed and implemented the Supervisor Assist product, giving team leads real-time visibility into ongoing customer interactions.
• Built a live monitoring dashboard showing all active calls for a team.
• Delivered real-time alerts on key conversational moments for supervisor intervention.
• Enabled secure live listening capabilities with minimal delay to monitor active agent-customer conversations.

Jul 2018 – Present · 7 yrs 11 mos

• As the founding engineer at Observe.AI, I played a pivotal role in architecting and building the core enterprise platform and data infrastructure that became the backbone for all AI-driven products. My work spanned platform architecture, data pipelines, and early product development — transforming the initial concept into a scalable enterprise-grade system.
• Enterprise Platform -
• Built a full-scale identity and access management system supporting users, CSMs, and external applications, leveraging OAuth and modern security standards.
• Enabled enterprise compliance and certifications (ISO, PCI, SOC2) by implementing secure policies, audit frameworks, and data protection mechanisms across the platform.
• Designed and developed account and user management systems, forming the foundation for multi-tenant operations and RBAC-driven authorization.
• Data Ingestion Pipeline -
• Architected and implemented a high-throughput data ingestion pipeline to acquire call recordings and metadata from diverse enterprise sources (SFTP, S3, APIs, etc.).
• Designed a modular, scalable processing framework to extract both ML-based and rule-based insights from millions of post-call recordings.
• Built the foundation that powers Observe.AI’s data-driven intelligence layer — ensuring reliability, security, and extensibility for large enterprise customers.

Jan 2018 – Jul 2018 · 6 mos