Pressable
A fortnight of bulk operations and a DDEV integration, aimed squarely at agencies
A side-by-side editorial comparison of incident.io and Heroku — release velocity, themes, recent moves, and the top alternatives to consider.
incident.io's AI now posts a root cause hypothesis before most responders have finished reading the alert.
Two threads run through the recent releases. One is steady on-call machinery — reassignable escalations, shift swapping, team-scoped private incidents, alert grouping without declaring an incident, richer Insights filtering. The other is the agent: reachable from anywhere in the web app since June, and now shipping Investigations to general availability, where Nexus posts a root cause hypothesis with supporting evidence into the incident channel within minutes of declaration.
Runtime bumps most weeks, but a new stack, a stack deprecation and an AI model cull underneath
Heroku's changelog is dominated by runtime maintenance - Node.js, .NET, Python, Ruby, Go and PHP updates land most weeks, often several in a day. Beneath that cadence the platform shipped real structural changes over May and June 2026: Heroku-26 reached general availability while Heroku-22 was deprecated, Heroku Connect opened to independent software vendors, a GitHub Enterprise Cloud integration arrived, Custom Trusted IPs came to Fir Private Spaces, and the CLI moved credential storage into system keychains. On the AI side, Heroku added Claude Opus 4.8 support in late May and then began deprecating select models from Managed Inference and Agents a week later.
Two threads run through the recent releases. One is steady on-call machinery — reassignable escalations, shift swapping, team-scoped private incidents, alert grouping without declaring an incident, richer Insights filtering. The other is the agent: reachable from anywhere in the web app since June, and now shipping Investigations to general availability, where Nexus posts a root cause hypothesis with supporting evidence into the incident channel within minutes of declaration.
incident.io is moving from coordinating the response to participating in it. The on-call and permissions work is what makes that safe at organization scale — private incidents scoped to teams, team-based permissions on alerts and escalations, workflow secrets and signed requests all landed before the agent got its most consequential job. The changelog also has a habit of burying real news in digest-titled entries, which is where MCP and the macOS app reached GA.
With a hypothesis and its evidence now posted automatically, the next step is acting on it — expect the agent's output to feed the workflow engine, and expect Investigations quality to become the thing incident.io is measured on.
Heroku's changelog is dominated by runtime maintenance - Node.js, .NET, Python, Ruby, Go and PHP updates land most weeks, often several in a day. Beneath that cadence the platform shipped real structural changes over May and June 2026: Heroku-26 reached general availability while Heroku-22 was deprecated, Heroku Connect opened to independent software vendors, a GitHub Enterprise Cloud integration arrived, Custom Trusted IPs came to Fir Private Spaces, and the CLI moved credential storage into system keychains. On the AI side, Heroku added Claude Opus 4.8 support in late May and then began deprecating select models from Managed Inference and Agents a week later.
Two things are happening at once. The platform layer is going through a generational turn - a new stack promoted, the previous one deprecated, and enterprise-facing controls (trusted IPs, keychain credentials, ISV access, GitHub Enterprise) filling in around it, which is the profile of a mature PaaS defending its installed base rather than chasing new workloads. The AI layer is behaving differently: adding a frontier model and pruning others within days points at a curated, opinionated inference catalogue rather than a broad marketplace. The relentless buildpack cadence is the constant underneath both.
Expect the Heroku-22 deprecation to progress toward an end-of-life date and further consolidation of the Managed Inference model list. The entries do not indicate what replaces the deprecated models, so whether the AI catalogue is narrowing or being re-based is not yet visible.
Other Infra & APIs products tracked by Sparkpulse, ranked by recent ship velocity. Tap any card for the full editorial trajectory or compare directly with incident.io.
A fortnight of bulk operations and a DDEV integration, aimed squarely at agencies
An LTS patch train quietly absorbing security fixes and multi-model AI switching
Jenkins is shedding weight and hardening deserialization on its weekly train
Honeycomb is turning Canvas from an investigation view into the console where the work actually happens
Cohere is shipping a model a month and retiring the old ones just as briskly.
Windsurf's own feature list has thinned out; what ships now is Devin arriving one surface at a time.
Other Infra & APIs products tracked by Sparkpulse, ranked by recent ship velocity. Tap any card for the full editorial trajectory or compare directly with Heroku.
HashiCorp is repositioning HCP Terraform as the accountability layer under AI agents that write infrastructure
Jenkins is shedding weight and hardening deserialization on its weekly train
Gravity Forms 3.0 makes accessible forms the default, not a checkbox someone forgets
Laravel runs two release trains at once, and 13.x is where anything interesting lands.
Bitwarden's server releases are all plumbing right now — flags cleared, billing untangled, invites hardened.
PlanetScale is building out Postgres while making every control reachable without the dashboard.
Latest ship moves from both products, interleaved chronologically. ⚡ = editorial spark.
They serve adjacent needs but don't currently overlap on shipped themes. incident.io is currently shipping more aggressively (velocity 6.3 vs 5.0), with 1 editorial sparks in the last 30 days against 0. See the at-a-glance table above for a side-by-side breakdown of velocity, recent sparks, and editorial themes.
Sparkpulse doesn't pick a winner — we score release velocity, not feature parity. incident.io is currently shipping more aggressively (velocity 6.3 vs 5.0), with 1 editorial sparks in the last 30 days against 0. For your specific use case, the alternatives sections above list other Infra & APIs products to evaluate alongside.
Top incident.io alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "incident.io alternatives" section above for the current picks, or visit /alternatives/incident-io for the full list with editorial commentary on each.
Top Heroku alternatives in Infra & APIs are ranked by recent ship velocity. Browse the "Heroku alternatives" section above for the current picks, or visit /alternatives/heroku for the full list with editorial commentary on each.