Jonathan Thinh

Led backend development of the Now Assist GenAI platform across ITSM and CSM domains, designing scalable microservices for the Generative AI Controller that orchestrates prompts, enforces guardrails, and integrates with multiple LLM providers (Now LLM Service, OpenAI, NVIDIA NIM). Architected retrieval pipelines for AI Search enabling contextual grounding of GenAI responses, including ingestion services to vectorize knowledge articles, service catalog entries, and incident history while applying ACL-based security trimming. Implemented backend APIs for the Now Assist Skill Kit, allowing developers to publish custom AI skills into workflows, and delivered REST/gRPC interfaces to enable platform integrations and customization of flow designed features. Containerized AI orchestration services and deployed on Kubernetes with autoscaling, reducing inference latency by approximately 30%. Built governance and observability features: backend logging pipelines (Elastic/CloudWatch), access contr

Backend engineer for Intuit GenOS, AI-powered financial platform on AWS enabling high-throughput data ingestion and model deployment across TurboTax, QuickBooks, Credit Karma, etc. Built scalable backend microservices (Java/Spring Boot, Node.js) supporting GenOS data integration and real-time analytics. Built and optimized big-data pipelines: batch jobs with Apache Spark on EMR and streaming workflows with Apache Flink on Kinesis to ingest and transform billions of financial records. Collaborated with data scientists to train and deploy ML models in Amazon SageMaker (NLP for customer support and forecasting models for tax/finance), exposing these via REST APIs as GenOS AI features. Migrated legacy data pipelines to containerized infrastructure (Docker, Kubernetes EKS) and automated provisioning with Terraform to improve reliability and reduce processing time. Implemented CI/CD with Jenkins and AWS CodePipeline for backend services and serverless functions (AWS Lambda). Improved data st

Core contributor to Teradata Vantage Platform; built backend services and AI/ML integrations powering predictive IoT, telecom, and retail use cases. Implemented scalable query engines, federated data pipelines, and multi-cloud deployments enabling petabyte-scale analytics. Integrated Aster Analytics engines (graph, ML, path analysis) into backend pipelines, enabling production-grade predictive analytics on IoT and log data. Produced in-database Python analytics capabilities and containerized execution layers for multi-language runtime integration. Designed and implemented Teradata IntelliSphere components (Data Lab, Unity, QueryGrid) for high-throughput data movement and near-real-time replication across analytic clusters. Developed integration connectors to enable federation across Teradata, Hadoop, Aster, and Oracle; optimized parallel joins and query performance. Created automated deployment tooling (Python, Terraform) for spinning up Teradata clusters on AWS and Azure. Built RESTfu