Platform Overview

What is StackFlow?

StackFlow is an agentic and cognitive AI portal designed for IT Service Management (ITSM) and infrastructure monitoring. Built natively on AWS, StackFlow combines enterprise-grade ITSM workflows with real-time AI assistance powered by Amazon Bedrock and Anthropic Claude models. Every incident, change, and service request is enriched with contextual AI intelligence drawn from your organization's knowledge base and configuration data.

⚙️ Minimum Requirements

AWS Account: 373544523367 with Lambda, Aurora, Neptune, ElastiCache, Bedrock, S3 enabled in us-east-1
Cognito User Pool: us-east-1_WKK1AVJ2m with at least one verified user and app client 570cnagpgoochn29a113du6jnt
Lambda: StackFlowAPI (nodejs22.x, arm64, 1792MB) deployed and healthy
KMS CMK: mrk-bd842691514c4d74a02992b8dc11fe16 enabled and accessible to all StackFlow IAM roles
CloudFront: Distribution E1UTZ9SVSR2WGV pointing to docs.stackflow-tech.com

Unlike traditional ITSM tools, StackFlow treats every workflow as an AI-first interaction. The platform learns from your resolved cases, continuously improving its recommendations through exemplar learning and semantic search. Whether you're triaging a P1 incident at 2 AM or building a complex change advisory board (CAB) workflow, StackFlow's cognitive engine is always active in the background.

Cloud-Native Architecture: StackFlow runs entirely on AWS infrastructure — Aurora PostgreSQL, Amazon Neptune, ElastiCache Redis, and AWS Lambda — giving you the reliability, scalability, and security of AWS with zero infrastructure management overhead.

Core Capabilities

StackFlow provides a unified platform for IT operations across five core domains:

Domain	Capability	AI Enhancement
ITSM	Incident, Problem, Change, Request lifecycle management	Auto-classification, suggested assignees, resolution drafts
CMDB	Configuration item discovery, Neptune knowledge graph	Blast radius analysis, dependency mapping
Knowledge Base	Article authoring, RAG-powered search	Auto-generated articles, semantic search via Bedrock KB
Cloud Management	Multi-cloud fleet health, FinOps, compliance	Anomaly detection, cost optimization recommendations
AI Workflows	Visual no-code automation designer	AI nodes, LLM-as-router, auto-remediation via SQS

Platform Architecture

StackFlow follows a serverless-first architecture on AWS. The primary entry point is an API Gateway REST API (uazcuhdus2) fronted by CloudFront, which routes all requests to the StackFlowAPI Lambda function running Node.js 22.x on ARM64 with 1792 MB memory and a 300-second timeout.

Data persistence uses Aurora PostgreSQL 16 for primary ITSM data and Aurora PostgreSQL 17 for requirements management. The Neptune knowledge graph stores CI relationships and is queried via Gremlin for blast radius and impact analysis. ElastiCache Redis provides sub-millisecond caching with TLS and authentication token enforcement.

CloudFront (E1UTZ9SVSR2WGV)
    │
    ▼
API Gateway (uazcuhdus2) → StackFlowAPI Lambda (nodejs22.x, arm64)
    │                              │
    │                    ┌─────────┼──────────┐
    │                    ▼         ▼          ▼
    │              Aurora PG    Neptune    ElastiCache
    │              (main/req)   Graph      Redis TLS
    │
    ▼
Bedrock KB (BXJGG7PIPS) ← OpenSearch Serverless ← Titan Embeddings v2

Key Differentiators

StackFlow is purpose-built for AI-native operations. The platform's semantic cache uses Redis to store vector-similar query results, reducing Bedrock API calls by up to 70% for frequently asked questions. The exemplar learning system captures resolved incidents with high satisfaction scores and uses them as few-shot examples for future similar cases.

All sensitive data is encrypted using AWS KMS Customer Managed Keys (CMK). The primary multi-region key mrk-bd842691514c4d74a02992b8dc11fe16 protects data at rest across Aurora, Neptune, and S3. Field-level encryption applies CMK-derived data keys to PII fields before they ever reach the database layer.

Tip: Start with the Quick Start Guide to have your first incident workflow running in under 15 minutes. No infrastructure setup required — StackFlow is fully hosted.

Getting Help

StackFlow provides multiple support channels. The AI Copilot sidebar (available on every page) can answer questions about your specific configuration in real time. The built-in Knowledge Base is searchable via natural language using the Bedrock RAG pipeline. For platform-level support, open a ticket at support.stackflow-tech.com or email support@stackflow-tech.com.

This documentation covers every aspect of the platform from initial authentication setup through advanced AI workflow configuration. Use the left sidebar to navigate by feature area, or press ⌘K to search across all documentation pages.

Quick Start Guide

Up and running in 15 minutes