Jan 2022 – Mar 2025 · 3 yrs 2 mos

• Led the AI Platform team (15+ engineers), delivering core services that generated ~80% of organisational revenue.
• Architected and optimised an OCR pipeline for unstructured data extraction, slashing enterprise underwriting times from days to minutes.
• Scaled real-time fraud detection to intercept over 10k attempts daily, and developed advanced deepfake detection and prevention capabilities.
• Engineered Aadhaar compliance workflows, enabling the compliant processing of over 100 million applications.
• Owned multiple internal 0→1 initiatives from ideation through product–market fit, executing iterative development cycles with lean teams.
• Maintained 99.97 %+ availability across product stacks via resilient architecture, automated monitoring and fault-tolerant deployments.
• Designed a GPU orchestration platform for inference and training—decoupling model deployment from engineering teams and handling over 5 million requests per month; growing at 20% MoM.
• Championed metrics-driven engineering by embedding observability at the core, with automated drift-detection, Evals at the core of training and annotation pipelines to track model variance as a leading indicator.
• Developed heuristic based fraud models combining camera and device-level signals with source-image analysis to detect deepfake attacks.
• Collaborated with enterprise clients to scale their onboarding solutions, increasing conversion rates by over 25%.

Mar 2021 – Feb 2022 · 11 mos

• Architected and horizontally scaled a face recognition platform, comprising face-search and metadata-retrieval pipelines—across a 50 million+ identity database, leveraging micro-services, FAISS vector search and ElasticSearch to achieve sub-100 ms query latency.
• Orchestrated a zero-downtime migration from on-premise GPU clusters to AWS, delivering a 3x reduction in end-to-end inference latency and a 5x increase in throughput.
• Engineered a face-based disbursement pipeline for a leading NBFC, integrating real-time biometric verification to cut loan disbursement times to under five minutes while reducing default rates.
• Containerised GPU inference services using Docker, Kubernetes and NVIDIA Triton Inference Server, implementing auto-scaling policies to boost throughput and system reliability under peak loads.
• Maintained 99.97% availability by deploying Grafana monitoring, automated scaling rules and resilient failover clusters within a lean engineering team.
• Upgraded to rotation-invariant face detection models, employing advanced data augmentation and optimised embedding networks to enhance both False Acceptance Rate (FAR) and False Rejection Rate (FRR).

Feb 2020 – Aug 2021 · 1 yr 6 mos

• Worked with 10+ NBFCs and Telecom companies in India & SEA to understand their business problems, design optimal workflows and do everything necessary to help them integrate our products, derive the most business value and solve business problems.
• Worked with the client's business stakeholders, tech stakeholders and developers to manage the entire process of planning, designing, integrating and deploying the solutions at scale. Everyday work included presentations, crucial conversations with clients, understanding and helping clients integrate our products in several different tech frameworks (Android, iOS, Java, JavaScript, .NET, React Native, C++, Python, etc) and platforms (Azure, AWS, Private Cloud).

Jan 2019 – Feb 2020 · 1 yr 1 mo · Bengaluru, Karnataka, India

• TheHouseMonk is an ML powered App based Customer Experience Platform for Property management, Facilities management and Real Estate Community management organisations. Adopted worldwide across 8+ countries and HQ'ed in Bengaluru; it is 'The Operating System' for every form of real estate management. Customers include Co-Living spaces, Co-Working offices, Rental Management companies, Commercial Tech. Parks, Gated Communities, Real estate builders and many more. A Young and energetic company embedding Technology into a legacy Industry.

Aug 2017 – Jan 2019 · 1 yr 5 mos · Bengaluru, Karnataka, India

• Architected a microservices-based web and mobile platform (Node.js/Express and React Native) deployed in Docker containers on AWS (ECS/Kubernetes) for elastic scaling and 99.9% uptime.
• Implemented real-time accounting and ticketing pipelines using WebSocket-driven APIs, PostgreSQL for ACID transactions, and Redis for low-latency caching.
• Collaborated with product, QA, and DevOps teams to define a domain-driven, multi-tenant SaaS architecture—leveraging isolated database schemas, OAuth2/JWT authentication, and GitHub Actions CI/CD for zero-downtime deployments.

Jan 2019 – Jan 2022 · 3 yrs · Bengaluru, Karnataka, India

• Led end-to-end development of multiple greenfield 0→1 products, from concept through launch.
• Architected and delivered the MVP for Sampingan—Indonesia’s hyperlocal, ad-hoc jobs marketplace, utilising a React Native front-end and Node.js services on AWS; scaled to 1 million+ installs, processed IDR 20 million+ in transactions and secured US$ 600 k in seed funding.

Jan 2016 – Aug 2017 · 1 yr 7 mos

• Collaborated with cross-functional stakeholders—both within the organisation and externally to architect and build horizontally scalable, high-volume web applications using micro-services, RESTful APIs and distributed data stores.
• Engineered end-to-end solutions for multi-dimensional client challenges, including lead lifecycle management, pre- and post-sales workflows, real-time tracking and analytics reporting.