Inherited Cloud Account Archaeology Tool For When the Engineer Who Built It Quits

dev tool venture scale •• multiple requests

Teams keep getting blindsided when their lead infra person leaves: undocumented services, design decisions only one brain knew, and outages that take 6+ hours because nobody knows where to look. AWS Resource Manager and CloudTrail show what's there but not why, what depends on what, or what's load-bearing in production.

builder note

Lead with the contractor angle — teams pay $100–500/hr to humans for exactly this. An AI that ingests CloudTrail+VPC flow logs+billing and outputs a 'here's what's load-bearing, here's what's orphaned' report wins on a per-account flat fee.

landscape (4 existing solutions)

Inventory tools exist but they answer 'what resources are here' not 'what would break if I deleted this'. The unmet need is a discovery+reasoning pass that produces a runbook from cold — call graphs from VPC flow logs, last-touched timestamps, cost concentration, and 'this looks like a bus-factor-1 component'.

AWS Resource Explorer + Tag Editor Lists resources but not call graphs, traffic relationships, or business-criticality. Multi-cloud blind.

Steampipe / CloudQuery Great query layer for cloud inventory, but you still have to write the questions — no opinionated 'what is load-bearing here?' output.

Lightlytics / Stream.security / Wiz Enterprise security-graph priced and scoped — overkill and over-budget for a 30-engineer shop trying to onboard a successor.

Backstage Service catalog only works if the predecessor populated it; doesn't auto-discover orphan resources or hidden dependencies.

sources (2)

reddit https://www.reddit.com/r/devops/comments/1suau0a/what_happen... "we only find out something exists when it breaks." 2026-04-24

reddit https://www.reddit.com/r/devops/comments/1suau0a/what_happen... "Hire expensive contractors at $100/hr to get you out of jail." 2026-04-24

cloud-archaeologyknowledge-managementawsbus-factorsuccession