Before they hit production. Built for engineers, priced for startups, works alongside Vanta and Drata.
Powered by agents I built. Shipped by someone who actually reads your Terraform.
It scored 35 out of 100. I wrote up every finding, the remediation, and what it cost to fix - Terraform diffs included. If you want to see what an honest AWS attack-path scan actually looks like, read the writeup.
Read the writeupVanta and Drata get you to SOC 2. Necessary - and they don't see your AWS attack surface. That's by design; they're audit infrastructure.
Wiz and Orca find the paths. Priced for security teams with the headcount and budget you don't have yet.
A scan a startup CTO can act on - alongside the compliance tool you already pay for. That's where I sit.
I run a read-only scan that maps your IAM graph and surfaces the paths an attacker could actually chain. Thirty minutes. You approve the IAM role before anything starts.
We go through every path together, prioritized by blast radius. You get the full markdown report with evidence citations and severity scored against your audit timeline.
I deliver the Terraform modules and IaC changes that close the gaps as PRs against your repo. Fixed scope, fixed price, no surprise invoices.
Built on direct AWS API calls, not heuristics. Five collectors today: IAM, CloudTrail, S3, Security Hub, Config. Read-only by construction. Methodology documented in the sample report.
Not sure which fits? Book a 30-minute scoping call and I'll tell you straight.
I built the tool I use. Four tiers of Claude agents plus a deterministic rules engine, read-only by construction. If you want to verify the methodology before trusting me with your environment, the sample report shows every step.
Hands-on with AWS, Terraform, and cloud security day-to-day. I'm not a consultant watching from the outside.
Go-based read-only collectors, deterministic scoring engine, agent-driven analysis. Built by one engineer who reads code, not status reports.
Zero mutating API calls exist in any collector. The tool can only observe your environment. I'll send you the IAM Terraform for the read-only role before the engagement starts so you can audit exactly what I can do.
The methodology is in the sample report. Every finding shows the underlying API call, the resource, and why it's a path. Show your work, not a black box.
I'm a security engineer based in the Bay Area, focused on AWS-native startups working toward audit-ready security. I've spent the last four years in IT and security, and I'm currently a security analyst. On the side I build the tools I'd want to use in that role - like kumo-assess, the analysis engine I use on every engagement.
I'd rather ship Terraform than slide decks. If you're past the "do we need SOC 2" conversation and want someone who can actually close the gaps, we should talk.
No, and that's the point. Vanta and Drata are how you prove SOC 2 to auditors - they weren't built to find graph-based AWS attack paths. I work alongside them: I find the IAM and infrastructure paths an attacker could chain; you keep using Vanta/Drata for evidence collection and policy management.
Speed and focus. A traditional readiness assessment takes 2 to 6 weeks and produces slide decks. My scan takes 30 minutes and produces Terraform. The agent-driven tool handles the data collection and analysis that consultants do manually, which means I can spend my time on the remediation work that actually matters.
Two ways. First, the sample report shows every step of the methodology end-to-end - I ran it on my own AWS account and wrote up every finding. Second, you create a read-only IAM role and I send you the Terraform for it before the engagement starts. The tool can only observe your environment; zero mutating API calls exist in any collector.
You create a read-only IAM role. The tool cannot modify anything in your environment - zero mutating API calls exist in any collector. I send you the Terraform for the role before the engagement starts so you can audit exactly what I can see.
AWS attack paths are framework-agnostic - they're real risk regardless of what you're certifying. Findings map to SOC 2 (CC6, CC7), PCI, ISO 27001, and CMMC controls. If you have a specific framework need, email me directly.
AWS today. GCP and Azure are planned for H2 2026.
Yes - I ran kumo-assess against my own AWS account and wrote up every finding. Read the case study.
Book a 30-minute scoping call. I'll ask about your environment, target audit date, and current state. You'll leave the call with a fixed price and a start date.
Book a 30-minute scoping call. I'll tell you straight where you stand and what it'll cost to close.
Book a scoping call