Amazon Quick + NVIDIA NeMo build supply-chain agent workflows from dashboards
Move from dashboard signals to ranked mitigation plans with natural language queries.
Amazon Quick and NVIDIA NeMo Agent Toolkit now offer a joint solution for building specialized agent workflows, starting with supply-chain risk. Amazon Quick acts as the business-user front door, providing a conversational workspace that connects to dashboards, structured data (via Amazon S3), and enterprise knowledge sources (Google Drive, SharePoint, Confluence, internal web). Users can also tap into over 100 pre-built action connectors (Outlook, Slack, Jira, Asana) and invoke agentic workflows through Model Context Protocol (MCP).
NVIDIA NeMo Agent Toolkit, an open-source, framework-agnostic library, powers the backend: registering tools, orchestrating the supply-chain investigation, capturing execution traces, and supporting evaluation/profiling for developers. In a typical workflow, an analyst asks diagnostic questions (e.g., 'What is behind this supplier delay?'), and Amazon Quick answers from dashboard context. When the analyst asks what to do next, Quick invokes the NeMo-backed MCP action. The workflow then investigates the disruption, validates recommendations, and returns a ranked mitigation plan. NeMo also provides telemetry, per-step latency, and evaluation results to tune the workflow—critical as teams move from simple chat assistants to multi-step agentic decision engines.
- Amazon Quick serves as the front door with dashboards, S3 knowledge sources, and 100+ action connectors (Outlook, Slack, Jira).
- NVIDIA NeMo Agent Toolkit orchestrates backend workflows using MCP, with telemetry and evaluation for tuning.
- Use case: supply-chain planners go from dashboard signal to ranked mitigation plan via natural language queries.
Why It Matters
Bridges dashboard visibility and decision execution, letting supply-chain teams automate complex investigation and recommendation workflows without adding headcount.