Senior Full Stack Engineer at Salt AI (2025-01 – 2026-03)
Built and scaled AI workflow orchestration systems enabling execution of complex LLM-driven pipelines and multi-step model workflows. Designed infrastructure to support agent-like execution patterns and collaborated on integrating LLM-based applications with production systems.
- Built and scaled AI workflow orchestration systems enabling execution of complex LLM-driven pipelines and multi-step model workflows.
- Designed infrastructure to support agent-like execution patterns, including dependency-aware task graphs, retries, and dynamic workflow composition.
- Collaborated on integrating LLM-based applications and automation systems, bridging model capabilities with production APIs and services.
- Improved reliability of AI pipelines through idempotent execution, retry policies, and observability, ensuring stable deployment of experimental and production models.
- Worked closely with product and AI teams to translate emerging AI capabilities into scalable, production-ready systems.
Senior Software Engineer at Ambience Health (2021-08 – 2024-12)
Improved platform reliability by implementing idempotent workflow execution and retry strategies. Designed transformer-based NLP systems for medical coding automation and built LLM-powered summarization pipelines for clinical data processing.
- Improved platform reliability by implementing idempotent workflow execution, retry strategies, and distributed tracing using Kubernetes, Redis queues, and Open Telemetry, enhancing system observability and reducing incident resolution time across AI workflow services.
- Designed and implemented transformer-based NLP systems for medical coding automation, leveraging deep learning models to extract structured insights from unstructured clinical data.
- Built LLM-powered summarization pipelines to generate structured patient insights, experimenting with prompt strategies and model configurations to improve accuracy and clinical relevance.
- Iterated on model performance through evaluation loops, prompt tuning, and data refinement, balancing accuracy, latency, and cost in production.
- Developed scalable systems that enabled continuous experimentation and deployment of AI models in healthcare workflows.
Senior Software Engineer at Zing Health (2019-05 – 2021-08)
Built and maintained full-stack features for the Member Portal using React, Node.js, and PostgreSQL. Designed REST APIs for Care Coordination platform and developed data pipelines for Population Health analytics.
- Built and maintained full-stack features for the Member Portal using React, Node.js, and PostgreSQL, enabling members to view benefits, check claims, access EOBS, and find providers, improving digital self-service adoption.
- Designed and implemented REST APIs and backend services for the Care Coordination & Case Management platform, enabling nurses and care managers to manage care plans, tasks, and social determinant workflows more efficiently.
- Developed data pipelines and analytics services using Python, SQL, and cloud infrastructure to support the Population Health platform, helping identify high-risk members and enabling proactive care interventions.
- Collaborated with product managers, clinicians, and operations teams to translate complex Medicare Advantage workflows into scalable platform services, improving system reliability and accelerating delivery of new member features.
- Optimized database queries and backend services, implemented monitoring and automated tests, and improved API performance, helping the platform support growing member enrollment while maintaining reliable access to critical healthcare data.
Software Engineer at Google (2015-02 – 2019-04)
Led backend development of scalable machine learning pipelines using TensorFlow. Designed production-grade ML infrastructure for training and deployment of speech recognition and vision models. Mentored junior engineers on distributed machine learning systems.
- Led backend development of scalable machine learning pipelines using TensorFlow and distributed data processing tools to support recommendation and computer vision models used across several internal Google AI services.
- Designed and implemented production-grade ML infrastructure for training and deployment of speech recognition and vision models, improving inference latency and reliability through optimized data pipelines and GPU-accelerated model execution.
- Collaborated with research scientists to productionize experimental ML models into scalable services, building reusable APIs and monitoring systems that enabled teams to deploy and iterate on AI features safely.
- Mentored junior engineers and contributed to architecture decisions for distributed model training workflows, gaining deep experience in large-scale machine learning systems, data engineering, and production AI reliability practices.
- Built and optimized machine learning training pipelines using TensorFlow, supporting large-scale experimentation in speech recognition and recommendation models while improving distributed training performance and dataset preprocessing workflows.
- Implemented backend services and data processing components that enabled researchers to train and evaluate computer vision models at scale using internal Google infrastructure and distributed storage systems.
Junior Software Engineer at Google (2013-06 – 2015-02)
Contributed to internal developer productivity tools used by engineering teams. Implemented features and performance optimizations for internal tooling supporting large codebases.
- Contributed to internal developer productivity tools used by engineering teams, improving build automation, debugging workflows, and development environments through backend services and UI improvements.
- Implemented features and performance optimizations for internal tooling that supported large codebases, learning best practices in scalable software architecture, code quality standards, and collaborative engineering workflows.
