Glasskube

What do they actually do

Glasskube builds open‑source tools that help software vendors deliver and operate their products inside customer‑controlled environments (on‑prem, VPC, air‑gapped, BYOC). Today they ship two projects: Distr, a "distribution control plane" with a Hub, agents, a registry, license controls and a UI/SDK, and the Glasskube package manager for installing and managing software on Kubernetes clusters. Distr can be self‑hosted via Docker Compose or Helm, and a hosted demo is available on distr.sh (Distr repo/docs, Distr docs, Glasskube package manager docs).

In practice, vendors connect their CI/CD (e.g., GitHub) to a Distr Hub, publish artifacts (OCI images, Helm charts, Terraform modules), and control which customers can access which versions. Customer environments install a lightweight agent or use standard clients to pull artifacts, run installs, send logs/metrics, and enable optional remote troubleshooting; vendors can push updates and automate via an API/SDK (Distr repo/docs).

Adoption is early: the code is public and actively developed, and early users are SaaS/AI vendors and DevOps teams needing to ship PoCs or production installs into customer infrastructure. Public customer references are limited; teams can self‑host or try hosted demos (YC profile, Distr repo).

Who are their target customer(s)

• SaaS/AI product teams selling on‑prem or air‑gapped installs: They don’t want to build and maintain a custom delivery system for each customer and struggle with slow, error‑prone integrations and license/version gating.
• Vendor DevOps teams who must deploy and update customer‑hosted instances: They spend time writing bespoke install scripts, dealing with environment differences, and manually debugging installs across many customer environments.
• Customer IT/platform teams running third‑party software in their VPC or on‑prem: They need predictable installs, clear audit trails, and a way to approve upgrades while maintaining control and security; they avoid unwanted automatic changes.
• Managed service providers / system integrators installing vendor software at multiple customers: They face repetitive, manual work across variable customer setups and need repeatable automation plus license/access controls to scale.
• Product/engineering managers who want distribution tied into CI/CD and product workflows: Lack of SDKs and integrations forces engineers to build one‑off automation, slowing releases and increasing maintenance.

How would they acquire their first 10, 50, and 100 customers

• First 10: Convert existing open‑source adopters: contact GitHub watchers and teams that installed the Distr Hub/agents to run time‑boxed, engineer‑assisted pilots in their VPC/on‑prem environments and turn 1–2 into paid references.
• First 50: Leverage early references and playbooks to target similar SaaS/AI vendors and MSPs: publish integration templates, host technical demos/webinars, list installers on Artifact Hub/Docker/GitHub, and sell short paid pilots or support subscriptions to de‑risk onboarding.
• First 100: Shift to repeatable motions: add an SDR/inside‑sales function to convert inbound OSS/trial signals, productize a hosted control‑plane with paid support/SLAs, build an MSP/SI partner channel, and close enterprise deals with security/compliance features and early case studies.

What is the rough total addressable market

Top-down context:

Starting from IDC’s $356B enterprise applications market in 2023, the relevant slice is the portion delivered into customer‑controlled environments; on‑prem/hybrid remains common in categories like ERP, and Kubernetes is widely used in production, supporting the need for vendor distribution tooling (IDC, ERP on‑prem share summary, CNCF survey).

Bottom-up calculation:

Illustrative bottom‑up: if 8,000–12,000 vendors globally offer on‑prem/BYOC options and could use a distribution control plane, and each spends $30k–$150k per year on tooling/hosted control planes/support, that implies roughly $240M–$1.8B in annual spend addressable by products like Distr.

Assumptions:

• 8k–12k global vendors offer customer‑hosted (on‑prem/BYOC/air‑gapped/VPC) options.
• Average annual spend per vendor on distribution/control plane tooling and support is $30k–$150k.
• A meaningful share prefers third‑party platforms over bespoke in‑house systems over the next few years.

Who are some of their notable competitors

• Replicated: Commercial vendor‑facing distribution platform for packaging, licensing, and delivering on‑prem/air‑gapped installs; long‑standing player with strong enterprise focus overlapping Distr’s goals.
• JFrog (Distribution / Artifactory): Artifact registry with distribution features for pushing binaries to edge or air‑gapped sites; competes on artifact delivery/sync even if less focused on vendor licensing and remote troubleshooting.
• SUSE Rancher (Fleet / Catalog): Multi‑cluster management with app catalogs and GitOps fleet features; an alternative when vendors integrate with customers’ cluster management instead of operating their own Hub/agent model.
• Portainer: Lightweight UI and agents for Docker/Kubernetes with app templates and an edge agent; smaller vendors/integrators can use it to install/manage apps across remote environments.
• Home‑built (OCI registries + Helm/CI + custom agents): Many vendors roll their own distribution using registries, Helm, CI/CD and scripts; Glasskube aims to replace this bespoke effort with an off‑the‑shelf control plane.