Contact us

Workflow

Efficient and responsible tasks handling is the basis of our service. In order to meet customer expectations, we have honed our skills and processes for over 10 years. For their better implementation, we have developed our own set of tools.

Scheduled tasks workflow

For scheduled tasks (those that do not require an urgent response) we use a workflow based on Scrumban methodology.

Planning and controlling of tasks is being carried in our internal services. Here are the most important of them:

  • Run — a Redmine-based ticket system (please check details below);
  • Nixon — our own service that provides task cards for the daily coordination of team activities.

In these systems, we have been focused on issues such as the deadlines control, automatic detection and indication of potential issues in tasks, the control of the response to changes by responsible persons, minimization of annoying notifications.

In addition, we have developed interfaces for every typical working case as well as a Slack bot for the submission of tasks.

Run

The Run service provides an accurate and timely implementation of planned tasks thanks to well-thought-out business processes, automation, and abnormalities detection.

  • Automation of processes ensures that all tasks come into the focus of the team at the right time.
  • Teamlead is responsible for the correctness of technical decisions.
  • To ensure that tasks are not fixed on individual employees, we immerse all engineers of the team in the specifics of projects and technologies used.

Urgent tasks workflow

Urgent tasks come from several sources:

  • mainly they originate from various monitoring systems that react to dangerous and potentially harmful changes in infrastructure or the application itself;
  • they may be initiated by our engineers who have noticed potentially dangerous changes;
  • they may be initiated by the customer.

Polk, our special internal service, gathers all of them and processes through a unified workflow based on Kanban. After that, they are handled by the duty engineer.

We strive to speed up the passage of urgent tasks through the pipeline by improving services and accumulating knowledge within the company.

Polk

The duty engineer processes all urgent tasks (including incidents) via Polk. This service is able to:

  • automatically prioritize tasks (obviously, the alerts of the production faulure are way more important than notifications of problems in some dev cluster);
  • collect information about the issue;
  • control the timely escalation of issues.

Polk tracks the time that incidents spend on each level of support. This helps engineers not to get stuck on any urgent yet difficult task. The rules of escalation take into account many parameters: where the incident occurred, what exactly went wrong, etc. Thus we make an escalating decision that is based on possible incident consequences, minimizing them along the way.