Jack Moes

Built and delivered full-stack AI capabilities for Zyrix Test Autopilot, Data Copilot, and TDM Copilot using React, Next.js, TypeScript, Python, FastAPI, Node.js, LangChain, LangGraph, and LlamaIndex. Developed AI agents for test planning, generation, execution, failure analysis, and retry decisions, and created MCP-compatible adapters to connect agents with test runners, data sources, APIs, and workflow services. Implemented RAG pipelines with Pinecone and semantic search, self-healing UI automation, API testing from OpenAPI/Postman assets, and data-quality workflows across PostgreSQL, MySQL, MongoDB, Snowflake, and Databricks. Deployed services with Docker, Kubernetes, Terraform, and CI/CD on AWS and Azure, while adding observability, security controls, audit trails, and human-review checkpoints.

Developed full-stack and machine-learning services that helped teams connect private enterprise data to foundation models. Built React, Next.js, and TypeScript workflows for data-source setup, ingestion monitoring, vector-store configuration, and retrieval testing. Created Python FastAPI and Java Spring Boot microservices for S3 ingestion, parsing, chunking, embeddings, metadata extraction, and vector indexing across more than 100,000 documents. Implemented RAG using Amazon Bedrock, Anthropic Claude, Titan Embeddings, OpenSearch Serverless, and Pinecone, improving answer relevance and retrieval quality. Added event-driven processing with Lambda and SQS, Redis caching, evaluation workflows, IAM/KMS security, CloudWatch monitoring, Docker, and CI/CD. Collaborated with product, science, security, and frontend teams to deliver production features across multiple release cycles.