Get Started
Evaluating an infrastructure or security company?
Get diligence from engineers who’ve run platforms at scale.
When reliability is the product, diligence has to go deeper
Diligence of infrastructure, developer tools, observability, and cybersecurity companies.
Infrastructure and security vendors sell trust. Their customers build on them, so an outage or breach doesn’t just cost revenue; it undermines the whole value proposition. We assess the platform’s reliability, its own security posture, and whether the engineering team can keep pace with a technically demanding customer base.
What we find most often
-
Differentiation that rests on execution, not assets.
Sophisticated engineering, but core techniques a well-funded competitor could reproduce. We assess what is truly proprietary: historical data, enrichment, and workflow integrations.
-
Multi-zone, but still single-region.
Infrastructure spread across zones while every service and data store depends on one cloud region, so a regional outage stops everything. We check for fault isolation (AKF calls these swim lanes): independent stacks so a failure in one customer group or feature can’t take down the rest.
-
Headroom that runs out mid-plan.
Capacity planned for today’s load. AKF’s rule of thumb is to design for 20x, implement for 3x, and deploy for about 1.5x; we test the platform against the growth plan, not current traffic.
-
Unclear quality ownership.
No defined QA function and thin end-to-end test automation, so defects reach technically demanding customers.
-
Tech debt and feature flags without governance.
No set cadence for paying down debt and no lifecycle for retiring flags, so complexity accumulates release by release.
-
Product decisions without discovery or KPIs.
Features built on internal conviction rather than validated customer need, with no metrics to show whether they delivered value.
-
Open-source exposure.
Licensing obligations, forked dependencies, or a community edition that cannibalizes paid revenue.