The News
Kedify, a Kubernetes autoscaling company built by core maintainers of the open-source KEDA project, announced expanded Insights capabilities at KubeCon North America 2026. The release combines CPU and memory rightsizing recommendations with autoscaling configuration analysis, including replica count tuning, trigger thresholds, and scaling behavior, into a single decision surface for platform teams. The expanded feature set also addresses performance at scale, with a distributed autoscaling architecture that Kedify claims outpaced native Kubernetes autoscaling in an AKS benchmark across large workload fleets.
Analyst Take
The problem Kedify is solving is real, and it’s getting more expensive
Kubernetes has become the default substrate for modern application delivery, but operational maturity inside those environments remains uneven. Platform teams routinely manage overprovisioned clusters not out of negligence, but because rightsizing resources and tuning autoscaling have historically been separate disciplines requiring separate tooling and separate expertise. The result is wasted capacity that’s difficult to quantify, harder to justify reducing, and almost impossible to optimize continuously as workload patterns shift.
Kedify’s move to unify resource sizing and autoscaling configuration into a single recommendation surface is the right architectural decision. The value isn’t just efficiency; it’s that the two problems are causally linked. Setting a minimum replica count too high defeats the purpose of rightsizing CPU requests. Setting it too low without adjusting scaling triggers creates latency spikes. Treating them as one coupled decision, rather than two separate tuning exercises, is how you actually reduce waste without sacrificing reliability.
Developer velocity as the hidden business case
The Tao Testing case study is telling. The customer’s headline benefit wasn’t a cost reduction percentage; it was that their team could “focus on new features rather than managing resource spikes.” That framing matters more than it might appear. According to ECI Research’s Google GovTech Survey Results, 47.2% of respondents selected “Developer velocity and ease of integration” as the factor carrying the greatest weight in their final technical selection process, ahead of both past performance and total cost of ownership. Kedify is, in effect, selling developer velocity wrapped in an infrastructure story, and that positioning has real purchasing traction.
This also connects to a broader pattern ECI Research has tracked. According to ECI Research’s Google GovTech Survey Results, 54.4% of respondents said infrastructure provisioning creates a moderate bottleneck where provisioning takes a few days and multiple tickets. That finding applies well beyond the public sector. Manual capacity management and reactive scaling are among the most consistent sources of friction between platform teams and the development teams they serve. Kedify’s Insights product attacks that friction by replacing judgment calls with evidence-backed recommendations that are reviewable before they’re applied.
Who wins, who should care, and what this means competitively
For ITDMs, the business case is straightforward: Kedify converts opaque infrastructure spending into a set of defensible, prioritized decisions with estimated cost impact attached. That’s the kind of artifact that earns sign-off in a quarterly business review. The ability to present resource reductions, capacity changes, and cost impact separately is a detail that matters in organizations where engineering and finance leaders are trying to speak the same language.
For developers and platform engineers, the more interesting capability is the distributed autoscaling architecture. Scaling KEDA itself across thousands of workloads has historically been a ceiling that organizations hit as their platforms matured. Kedify’s approach of distributing autoscaling work across independent groups while preserving familiar KEDA configurations is an architectural answer to a real operational constraint. Teams running burst-driven APIs, queue processors, or AI inference workloads, where cold-start latency is a production concern, have the most to gain.
Competitively, Kedify’s founding team credentials matter here. Being built by KEDA’s core maintainers isn’t just a marketing point; it means the product decisions are informed by the deepest possible understanding of where the open-source project’s limits are. That’s a meaningful moat in a market where most autoscaling tooling is built on top of KEDA, not by the people who built KEDA.
Looking Ahead
Kedify is positioned in a market that will expand as Kubernetes adoption deepens and AI inference workloads become a larger share of cluster traffic. AI inference is a particularly demanding autoscaling target: traffic is bursty, latency requirements are strict, and model loading times make cold starts expensive. Kedify’s combination of event-driven and predictive scaling is well-suited to that problem, and we expect the company to make inference workload optimization a more explicit part of its product narrative over the next 12–18 months.
The broader trajectory here is consolidation of platform observability and optimization into fewer, more opinionated tools. Teams that today manage rightsizing, autoscaling, and cost allocation through separate products will face increasing pressure to simplify their platform engineering stack. Kedify’s decision to combine Insights with autoscaling management puts it on the right side of that consolidation. The question for the company is whether it can grow from a specialist autoscaling tool into a broader platform efficiency layer before larger observability and cloud cost management vendors make the same move from the other direction.
Stay Ahead of Application Development Trends
Get weekly analyst insights, research notes, event coverage, and AppDevANGLE updates delivered directly to your inbox.
Subscribe for Weekly Insights
Join technology leaders, practitioners, and GTM teams following the trends shaping modern software delivery.
Looking for deeper research access?
Explore ECI Research reports, survey insights, and market analysis through the ECI Research Portal.
