Skip to content

Platform engineering

An internal platform your engineers actually use

Self-service environments, operational tooling and internal portals — built by the engineers who take the on-call shift themselves.

The problem

Every request goes through a person

A test environment needs a ticket and waits two days. Access is granted by the one engineer who knows every system. The knowledge is in people's heads, the procedure is in a chat thread, and every new hire asks exactly the questions the last one did.

  • Routine work still needs a ticket
  • Access depends on one person
  • Each team does it its own way
  • A tool was bought and nobody uses it

What we deliver

An internal platform your engineers actually use

Service catalogue

What an engineer can request themselves, what needs approval, and what stays manual.

Self-service environments

Templates and automated creation — a test environment in minutes rather than days.

Internal portal

One place where system state, procedures and tools sit together.

Access model

Roles, approval steps and an audit trail — who did what, and when.

Runbook automation

A repeated procedure becomes a button, with logging and a result check.

Adoption

Training and documentation — a platform nobody uses is a cost.

Architecture

Architecture

  1. 01RequestThe engineer picks from the catalogue
    • Service catalogue
    • Golden paths
  2. 02AuthoriseRole and approval step
    • OIDC
    • RBAC
    • MFA
  3. 03ProvisionA template rather than manual steps
    • Terraform
    • Ansible
    • cloud-init
  4. 04PortalState and procedures in one place
    • Next.js
    • REST API
    • OpenPath
  5. 05AuditWho did what, and when
    • Audit log
    • Graylog
  6. 06AdoptionMeasuring actual use, and training
The path of a request — from ticket to button, with an audit trail

Capabilities

Capabilities

Every item is marked: verified production experience, or engineering capability.

A NOC console with live monitoring

Proven

OpenPath — our own console with live path monitoring, in daily use on the group's network.

  • OpenPath
  • TypeScript
  • PostgreSQL

A diagnostics tool platform

Proven

noc.com.ge — 33 network tools in one place, publicly available and the daily instrument of our own NOC.

  • noc.com.ge
  • PHP
  • REST

A control panel with billing and provisioning

Proven

The hosting platform's own panel: plans, balance and server creation at API level — a working product, not a prototype.

  • Next.js
  • PostgreSQL
  • REST API

Templates and automated environments

Capability
  • cloud-init
  • Terraform
  • Ansible

SSO and role-based access

Capability
  • OIDC
  • RBAC
  • Audit log

Internal developer portal

Capability
  • Service catalogue
  • Golden paths

Process

Process

  1. 01

    Demand map

    Which requests repeat most often.

  2. 02

    The first golden path

    One scenario automated end to end.

  3. 03

    Portal

    The interface where that scenario is available.

  4. 04

    Extend

    The next scenarios, on the same template.

  5. 05

    Access and audit

    Roles, approvals and the trail.

  6. 06

    Adoption

    Training, documentation and measuring actual use.

Technology stack

Technology stack

Platform
TypeScriptNext.jsPHPPostgreSQL
Automation
AnsibleTerraformcloud-initREST API
Access
OIDCRBACMFAAudit log
Visibility
GrafanaPrometheusGraylog
Our own tools
OpenPathnoc.com.ge

Engagement model

Engagement model

Project

A one-off scope: audit, migration or implementation with a fixed outcome.

Retainer

Monthly engineering hours — specialist access on demand.

Co-managed

NetWizard and your in-house team together, with split responsibility.

Use cases

Use cases

A test environment in two days

The developer waits while the infrastructure team is busy. We build the template and self-service creation — the request moves from a ticket to a button.

The off-the-shelf product does not answer the question

The view you need exists in no tool. We write our own — the way OpenPath and noc.com.ge were written for our own NOC.

The knowledge is in people's heads

Procedures are found by searching a chat. We turn them into runbooks, then into automated steps with logging.

FAQ

FAQ

Isn't this only for large companies?

No. Even in a team of five, automating three repeated procedures frees a day a week. We start with the one that repeats most.

Wouldn't an off-the-shelf product be better?

If it is better, we choose it. We write our own when the ready-made tool does not answer the question, or when integrating it costs more than building it.

Who owns the code?

You do. The tooling we build stays in your repository with documentation and a handover — we can stay on support, but that is a choice, not a dependency.

Tell us about your infrastructure