StatusGator has a peculiar obligation: it needs to remain available when other services disappear. Its customers arrive looking for explanations of somebody else’s outage. A blank page would be an unusually concise answer. For years, the company depended on Cloudflare. Then a November outage sharpened the question its co-founder Colin Bartlett had already been asking: what happens when the network protecting your application becomes the reason nobody can reach it?
- The proposition: operate several delivery networks through one control layer.
- The evidence: customer stories describe lower bills, faster downloads and survival during provider incidents.
- The useful lesson: test alternatives while the existing system still works.
“We need to be up when everyone is down,” Bartlett says in IO River’s customer account. StatusGator adopted IO River’s multi-CDN platform and Check Point WAF. Soon after onboarding, the account says, a CloudFront outage arrived and the service stayed available. The buying decision had become a production test.
01 / The network above the networks
A content delivery network, or CDN, keeps content close to users and helps shield the application behind it. Choosing one brings welcome simplicity. Choosing several brings another collection of interfaces, configurations and decisions. Someone must keep their behavior aligned and decide where requests should go. Otherwise, redundancy becomes a second job.
IO River sells that missing coordination. It connects to providers through their APIs, brings configuration and monitoring into one console, and steers traffic according to availability, performance and cost. Its control plane manages the decisions without becoming another proxy through which all content must travel. Customers can bring existing CDN accounts; a managed Virtual CDN offering also combines underlying networks into one service.
available
degraded
available
The distinction matters in a market where Akamai, Cloudflare, Fastly and CloudFront can be alternatives to IO River’s proposition and suppliers underneath it. IO River’s pitch is to make those networks easier to combine. Security policies, certificates and edge functions also need consistency when the chosen route changes.
02 / The engineer who said no
Edward Tsinovoi’s account of meeting Michael Hakimi begins with a rejected job offer. Tsinovoi was interviewing him for a principal engineering role; Hakimi declined. Roughly twelve years later, the two were co-founders of IO River. A refusal, apparently, need not spoil a promising professional relationship.
Founded in 2022 by the former Akamai engineers, the company draws on their experience building CDN and edge systems. Tsinovoi’s launch essay describes recruiting former colleagues and engineers with experience at Google and Dell, then adding commercial expertise from CDN and security businesses. The public account emphasizes experienced engineering and close attention to customer problems.
03 / Sometimes the supplier leaves first
For Minute.ly, the trigger was more abrupt. StackPath announced it was shutting down its CDN. Minute.ly’s video technology depended on advanced configurations, custom capabilities and API integration. Finding another network was only part of the assignment. The replacement had to preserve the behavior its product already expected.
IO River helped assess vendors, supplied capabilities missing from the selected replacement, and supported a gradual transfer of production traffic. Minute.ly compared the old and new delivery behavior throughout. The published case study estimates three months of DevOps effort saved and reports a 35% improvement in worldwide median time to first byte. These are customer-specific results, not a migration timetable for everyone.
04 / A very small clock, a very large difference
Shuffll’s problem was less dramatic and more revealing. The video platform used Azure and Fastly, and cared about how quickly static files finished downloading. IO River’s monitoring exposed a short cache lifetime: cached content was cleared every three minutes. Azure struggled with that pattern more than Fastly did.
The team enabled a dynamic traffic rule to select the better-performing CDN for each user. IO River’s account reports a 79% improvement in worldwide median time to last byte and 66% in the United States. Time to last byte measures completion; time to first byte measures when a response begins. They answer different questions. A respectable benchmark can still miss the thing your application needs.
“We need to be up when everyone is down.”Colin Bartlett · StatusGator co-founder
05 / The bill is part of the architecture
Minute Media, whose brands include The Players’ Tribune and FanSided, had a growing delivery bill. Its original provider offered reliable service, but adding networks traditionally meant expensive operational work. According to the case study, a proof of concept took a few days, followed by gradual adoption. Reported results: 30% lower CDN costs and 50% better worldwide time to first byte.
IO River charges for orchestration. Its AWS Marketplace listing shows monthly tiers from free for delivery up to 30TB to $250 for 250TB and $1,500 for 5PB. Those are orchestration fees, rather than the entire delivery bill. Provider charges remain part of the calculation. Cost-based steering considers performance and contractual commitments, so a cheaper route on paper may not be the economical route under your agreement.
06 / Even redundancy has dependencies
IO River itself had a dependency to reconsider. IBM’s March 2026 case study describes constraints from relying on one DNS provider. Integrating NS1 Connect helped it translate complex policies into DNS decisions, including routing by autonomous system number, which identifies a network. Multi-provider DNS support reduced reliance on one supplier. The coordinator needed resilience too.
The company announced a $20 million Series A in January 2026, led by Venture Guides and New Era, after its $5.4 million seed round. Its ambition extends to separating security and compute services from delivery infrastructure. At the Series A announcement, it reported managing more than 200 petabytes of traffic monthly.
The method worth copying is modest: compare providers against your workload, move traffic gradually, watch errors and speed, and retain a route back. Several CDNs cannot rescue a failed origin or a shared bad configuration. Nor does every small application need this machinery. IO River becomes interesting when the delivery bill, the audience or the consequences of an outage justify keeping more than one door open.