Infrastructure and networks

33 use cases, 11 specialists

DevOps engineer

  • Setting up CI/CD with AI

    Writes a build, test, and deployment pipeline for GitLab CI or GitHub Actions with a security check and automatic rollback.

  • AI agent writes Terraform: infrastructure as code

    Describes servers, networks, and databases in Terraform files so that the environment can be brought up again with a single command and all changes are visible.

  • Docker and Kubernetes: AI assistant for zero-downtime deployment

    Writes a Dockerfile, a docker-compose file, and Kubernetes manifests with health checks, autoscaling, and zero-downtime deployment.

Site reliability engineer (SRE)

  • Availability targets (SLOs) and error budgets: AI assistant for SREs

    Suggests availability targets based on the service's metrics, calculates the error budget and its burn rate, and draws up alerting rules.

  • Prometheus and Grafana monitoring with AI

    Prepares metrics for the four golden signals (latency, traffic, errors, saturation), alerting rules, and Grafana dashboards.

  • Failure drills (chaos engineering): AI prepares the scenario

    Draws up a drill plan with a hypothesis, blast radius limits, steps, and stop conditions, and afterward a report on the weak points found.

System administrator

  • AI for system administrators: troubleshooting failures

    Finds the likely cause of a failure on a Linux or Windows server from logs and command output, and suggests commands to fix it along with a rollback procedure.

  • Server backups: AI prepares the plan and scripts

    Draws up a backup scheme with a schedule, retention periods, and restore testing, and writes the scripts for it.

  • Securing a Linux server: an AI check of security settings

    Checks access rights, updates, open ports, and logs using command output, and produces a list of fixes with commands.

Incident commander

  • Blameless incident postmortem: AI assistant

    Writes a postmortem from the timeline in the chat and logs: impact, causes using the 5 Whys method, and action items with owners and deadlines.

  • Incident response procedure with AI

    Draws up severity levels, roles, an escalation procedure, and status message templates for the team, management, and customers.

  • On-call schedule and runbooks for failures: AI agent

    Draws up an on-call schedule that does not overload people and step-by-step runbooks for known failures with result checks.

Network engineer

  • Cisco, Juniper, and Palo Alto configuration with AI

    Writes a router, switch, or firewall configuration for the task, taking the platform into account, with verification commands.

  • AI assistant for network troubleshooting from command output

    Analyzes show command output, routing tables, and interface counters, and names the likely cause and the next commands to run as a check.

  • Network change plan with rollback: prepared by AI

    Draws up a work plan: pre-checks, steps, verification commands to run afterward, and exact rollback commands.

Network engineer (Huawei, H3C, Eltex)

  • Configuring Huawei and Eltex switches with AI

    Writes a VLAN, OSPF, or BGP configuration for Huawei, H3C, or Eltex to match your topology, with display commands for verification.

  • Replacing Cisco with Huawei or Eltex: AI converts the configuration

    Converts a Cisco configuration into Huawei, H3C, or Eltex syntax and flags features that have no direct equivalent.

  • Russian FSTEC network requirements: AI assistant for network engineers

    Compares the network diagram and settings with the requirements of FSTEC of Russia (the Federal Service for Technical and Export Control) for significant critical information infrastructure (CII) facilities, and compiles a list of nonconformities for a specialist to review.

Database optimizer

  • Speed up a slow SQL query: AI analyzes the execution plan

    Analyzes EXPLAIN ANALYZE output, finds the bottleneck, and suggests a rewritten query and missing indexes.

  • PostgreSQL indexes: which to add and which to drop, with AI

    Selects indexes (composite, partial, GIN) based on the schema and frequent queries, checks that foreign keys are indexed, and finds redundant indexes that only slow down writes.

  • Tuning PostgreSQL and MySQL for load: AI assistant

    Suggests memory, connection, and PgBouncer pool settings based on server parameters and query statistics, and explains each change.

Database reliability engineer

  • PostgreSQL backups with restore testing: AI assistant

    Draws up a backup scheme with point-in-time recovery, writes a script that regularly tests the backups by restoring them, and calculates whether you stay within the acceptable data loss.

  • Database replication and failover: planned by an AI agent

    Designs a replication and automatic failover setup, replica lag monitoring, and a step-by-step plan for failover drills.

  • Zero-downtime database migration with AI

    Breaks a table change into safe "expand, migrate, contract" steps, checks which commands will take a lock, and writes a rollback plan.

Cloud cost engineer

  • How to cut cloud costs: AI finds the waste

    Finds idle machines, forgotten disks and snapshots, and oversized resources in a billing export, and calculates the savings for each fix.

  • Allocating cloud costs by project with AI

    Suggests a tagging scheme and allocates the cloud bill across teams, services, and environments, including the split of shared costs.

  • Cloud budget forecast and cost per customer: AI report

    Builds a report from exports with costs per customer or per request, a forecast to the end of the month, and flags for unusual spikes.

API cost watchdog

  • How much AI model requests cost: AI calculates your API spend

    Calculates the cost and response time of each call and scenario from request logs and shows where money is being wasted.

  • API overspend protection: limits and emergency shutoff with AI

    Writes a wrapper for API calls with a timeout, a retry limit, a spending cap, and automatic shutoff when errors spike.

  • Choosing a cheaper model without losing quality: AI comparison

    Runs the same tasks through backup models in the background, scores the answers on a scale set in advance, and suggests which share of requests to switch and where.

IT service manager (ITIL)

  • ITIL-based IT service catalog: AI assistant for IT departments

    Draws up a catalog of the IT department's services with owners, the ordering procedure, and service level agreements in a Word document.

  • AI for incident classification and prioritization

    Draws up a priority matrix by impact and urgency with escalation rules, and flags misclassified tickets in a ticket export.

  • SLA report from a ticket export with AI

    Calculates deadline compliance for each service from a ticketing system export, finds breaches, and prepares a report in Excel.