This quarter’s RepliCon brought together customer stories, product innovation, and best practices from the frontlines of enterprise software delivery. The common theme across all four talks was removing uncertainty—whether in installations, onboarding, real-world deployments, or distribution models. Our speakers shared how Replicated helps vendors deliver secure, consistent, and repeatable experiences to enterprise customers.
Here are the highlights, organized for easy browsing and exploration.
Why Better Installs Start Before Installation
Speaker: Evans Mungai, Senior Customer Reliability Engineer
Session: Preparing for Installs & Reducing Failures
Evans shared lessons from years of troubleshooting installs in the wild, where environments range from locked-down to loosely managed—and surprises lurk at every step. Survey results revealed the challenge: only 1 in 8 vendors believe they deliver a “delightful” install, and 20% of installs exceed 30 minutes, with some stretching into weeks.
Common blockers:
- Security tools like Microsoft Defender, Lumio, and Symantec that silently break installs.
- Misconfigured infrastructure, from load balancers to under-provisioned disks.
- Automation conflicts, such as Ansible reverting configs mid-install.
Recommendations:
- Provide role-specific prerequisite guidance so admins, infra teams, and app installers each see only what’s relevant.
- Use Preflight to codify checks and provide a traffic-light validation experience.
- Require evidence-based validation—preflight bundles as proof before installation day.
- Feed lessons back into centralized catalogs or product improvements.
- Document organizational prerequisites in Enterprise Portal, customized by install type.
Takeaway: Preparing customers properly—through clear guidance and codified checks—transforms installs from multi-day marathons into predictable, first-attempt successes.
Portal Kombat: Defeating Bad User Onboarding Cannot Be Left to Chance
Speaker: Chuck D’Antonio, Director of Solutions Engineering
Session: Portal Kombat: Defeating Bad User Onboarding Cannot Be Left to Chance
Chuck gamified onboarding with a Mortal Kombat metaphor: out-of-the-box Enterprise Portal already improves on “tarball over email,” but real wins come from deliberate customization.
Basic combos (the must-dos):
- Custom domain (e.g.,
enterprise.yourdomain.com) - Branded UI (logos, colors, company identity)
- Branded, DKIM-verified emails
Power moves (to elevate trust):
- Tailored pre-/post-install instructions per install method
- Optional Security tab to proactively share SBOMs and scan results
- Trial signups with flexible license types
Fatality (full control):
- Go headless with Enterprise Portal APIs, embedding onboarding flows into your own portals (Zendesk, Salesforce, or custom-built).
Chuck’s demo of “Slacker News” illustrated the leap from generic portal to fully branded experience—proving that thoughtful customization builds trust from the first interaction.
Takeaway: Onboarding is your customer’s first impression. Branding, transparency, and APIs ensure it’s a flawless victory.
Pixee’s Approach to Easier Enterprise Deployments
Speaker: Brian Doyle, Staff Engineer at Pixee
Session: Pixee’s Approach to Easier Enterprise Deployments
Brian described why Pixee adopted Replicated: to handle variability, strict security requirements, and operational complexity when delivering AI-powered security software into enterprises.
Capabilities that made the difference:
- Embedded Cluster: Fast installs (as quick as 30 minutes), with built-in proxy and PKI support.
- Headless installs: Scriptable automation for customers who rebuild dev/test daily or deploy at scale.
- Support Bundles & Preflight: Proactive validation and ongoing visibility.
- Licensing & Channels: Streamlined trials, dev/test, and production tiers.
- Compatibility Matrix: CI-integrated testing across EKS, AKS, GKE, and VM-based deployments.
Wisdom to share:
- Codify preflights early and require them as gates.
- Maintain environment-specific documentation.
- Monitor integrations with support bundles.
- Mirror complex environments in lab settings for regression testing.
- Plan for both Helm and Embedded Cluster installs.
- Engage as a design partner (e.g., with Embedded Cluster v3) to shape roadmap priorities.
Takeaway: Replicated gave Pixee the flexibility to handle “every customer is different” with confidence and repeatability.
Distributing Your Replicated K8s Appliance as a Virtual Machine – Azure Edition
Speaker: Martez Killens, Senior Solutions Engineer
Session: Distributing Your Replicated K8s Appliance as a Virtual Machine – Azure Edition
Martez built on Chuck’s earlier VMware OVA template demo by showing how to package Replicated-powered appliances as Azure VM images. The approach removes OS-level variability and provides a single-file delivery model for customers.
How it works:
- Vendor side: Use Packer to bake assets into Ubuntu-based images, ARM templates for networking and security, and Cloud-init scripts for installs. Publish via Azure image galleries for global distribution.
- Customer side: Deploy VM from the gallery, launch Embedded Cluster + Replicated install automatically, and access the admin console with minimal setup.
Benefits:
- Eliminates OS drift and file system surprises.
- Reduces security/policy misconfigurations.
- Provides end-to-end click-ops installs—from OS to Kubernetes to app.
- Scales globally via Azure’s distribution model.
Takeaway: Extending Embedded Cluster into prepackaged Azure VMs smooths the entire delivery chain, delivering consistency where it matters most.
What’s Next?
Replicated continues to evolve into a comprehensive platform for secure, repeatable, and enterprise-friendly commercial software delivery. If any of these features sparked your interest, reach out to your account team—or come chat with us in Slack.