Skip to content

index.md

Latest commit

 

History

History
68 lines (44 loc) · 4.1 KB

index.md

File metadata and controls

68 lines (44 loc) · 4.1 KB
Raw

Deployment

The Deep Researcher Agent blueprint supports multiple deployment methods. Choose the one that best fits your environment and operational requirements.

Method Best For Prerequisites
Docker Compose Local development, team demos, single-node deployments Docker Engine, Docker Compose v2
Kubernetes (Helm) Multi-node clusters, production Kubernetes cluster, Helm v3.x
Manual (no containers) Development and debugging Python 3.11--3.13, system dependencies (refer to Installation)

Published Release Artifacts

Deep Researcher Agent v2.2.0 is published on NVIDIA NGC. Use the exact versioned references below for release deployments.

Artifact Type Versioned reference
aiq-agent Container image nvcr.io/nvidia/blueprint/aiq-agent:2.2.0
aiq-frontend Container image nvcr.io/nvidia/blueprint/aiq-frontend:2.2.0
aiq2-web Helm chart nvidia/blueprint/aiq2-web:2.2.0

Architecture Overview

All containerized deployments run the same three services:

  • Backend (deep-researcher-agent) -- FastAPI server with an embedded Dask scheduler and worker for background job processing.
  • Frontend (deep-researcher-blueprint-ui) -- Next.js web UI that communicates with the backend API.
  • Database (postgres) -- PostgreSQL instance for async job storage, LangGraph checkpoints, and document summaries.

Deployment Guides

  • Docker Compose -- Full Docker Compose reference covering environment setup, the standard LlamaIndex stack, Foundational RAG (FRAG) integration, database configuration, and troubleshooting.

  • Kubernetes (Helm) -- Helm chart deployment for Kubernetes clusters, including NGC image pull secrets, configuration switching, FRAG integration, and troubleshooting.

  • Amazon OpenSearch Serverless -- EKS and OpenSearch Serverless deployment notes for the built-in OpenSearch knowledge backend.

  • Docker Build System -- Multi-stage Dockerfile architecture, build targets (dev vs. release), base images, and startup scripts (entrypoint.py and start_web.py).

  • Authentication -- Enable OAuth/OIDC sign-in, configure backend JWT validation, and use Deep Researcher Agent user tokens in tools and MCP pass-through integrations.

  • Async Job Content Encryption -- Configure encryption at rest for async final reports and selected artifact event content, including Vault Transit and static-key modes.

  • Observability -- NeMo Relay logging, ATOF traces, Phoenix OTEL export, redaction, and cost data.

  • Production Considerations -- Guidance on managed databases, horizontal scaling, security hardening, monitoring, and resource requirements.

  • OpenShell -- Optional policy-bound execution for generated deep-research code, including supported platforms, authenticated gateway ownership, policy/config pairing, deterministic live acceptance, and safe cleanup.

Quick Start

For the fastest path to a running stack:

# 1. Configure environment
cp deploy/.env.example deploy/.env
# Edit deploy/.env with your API keys

# 2. Start services
cd deploy/compose
docker compose --env-file ../.env -f docker-compose.yaml up -d --build

Open http://localhost:3000 to access the web UI.