Owned the end-to-end (0ā1) development of AI agents across logistics workflows (e.g., Order Agent, Load Agent, Fuel Station Recommendations, Payroll Agent), building React (JavaScript/TypeScript) front-end dashboards and Node.js backend services with LLM prompt engineering, schema-validated JSON extraction, and REST API integrations.
Designed and built admin and ops tooling using React/TypeScript dashboards and JSON-based metrics to monitor and manage AI agent performance, accuracy, extraction quality, and throughput, aligned with Manage AI Agent capabilities (real-time tracking, escalation/engagement visibility, feedback loops).
Developed AI-driven logistics agents using LLMs and RAG, integrating React (JavaScript/TypeScript) UIs, Node.js services, prompt engineering, schema-validated JSON extraction, and REST APIs to convert PDFs/files into structured orders within TMS/CRM systems.
Collaborated with Operations, Product, and Engineering teams to design and ship AI automation using React (JavaScript/TypeScript) front-ends, Node.js backend services, and API-driven workflows, reducing manual data entry and accelerating order creation across supply-chain operations.
Improved LLM prompting and validation pipelines using JSON schemas, structured outputs, guardrails, and retry/fallback logic, increasing extraction consistency and reliability across diverse document formats.
Ensured scalability, maintainability, and front-end performance across enterprise dashboards through modular React architecture, TypeScript, and API-efficient component design.
Implemented and maintained front-end test coverage using Jest and React Testing Library, validating critical user flows, API integrations, and AI-driven workflows with mocked LLM and API responses.