Dmitriy Khodak

DevOps and Cloud Engineer, AI-driven DevOps

Location
Kazakhstan, remote
Experience
18 years IT, 5+ cloud
Languages
Russian native, English C1
Current
Contract, Cyprus company
Open to
Relocation and business trips

Summary

DevOps and Cloud Engineer with 18 years in IT and 5+ years owning the full delivery stack (IaC, CI/CD, containers, observability) for a production platform serving 10,000+ users across 500+ Linux hosts.

I migrated a live AWS estate to Terraform with no downtime, run pipelines on GitHub Actions and Jenkins, and write the Python/FastAPI services that connect systems. Strong on the operational side as well: 2nd/3rd-line incident response, root-cause analysis, secure engineering practice and ITIL v4 change management.

On the AI side I run a RAG course assistant on Amazon Bedrock in production and apply LLMs to log correlation, first-pass RCA, toil scripts and change pre-review, under the same review and audit trail as any human-authored change.

Experience

Senior IT Engineer, OneLearn Global

to present

Cyprus, remote. Enterprise e-learning for the maritime industry: an AWS-hosted platform plus integrations with external HR systems and simulator vendors.

  • Migrated live AWS production infrastructure (VPC, subnets, EC2, RDS MySQL, Lightsail) to Terraform with remote S3 state using import blocks. Zero downtime, 60% faster deployments, and version-controlled IaC in place of undocumented manual provisioning.
  • Codified the production Moodle/IOMAD stack in Terraform: VPC across two availability zones, EC2 with IMDSv2 enforced, RDS MySQL in private subnets with deletion protection, and security groups. Added a bootstrap root (GitHub OIDC, CI role, SSM parameters), a disposable dev environment built from the golden AMI, and a reusable Lightsail module for the Node.js and AI apps.
  • Own the CI/CD platform end to end across Jenkins and GitHub Actions: build, test, image push and automated deployment with rollback-ready pipelines.
  • Built and operate a Python/FastAPI integration hub connecting the enterprise platform to a third-party RDP-based simulator: LTI 1.3 OIDC launch (RS256 JWT, JWKS, AGS grade passback), an HMAC-signed legacy launch flow, and xAPI score reporting that is idempotent against re-grades. Every external call has exponential-backoff retries and dead-letter handling, so the workflow keeps running through vendor outages.
  • Built and run in production an AI course assistant on Amazon Bedrock (Titan embeddings, Cohere Rerank, Claude). It answers learners only from course content, cites the source slides and declines out-of-scope questions. Hosted on Terraform-provisioned Lightsail with a Bedrock-only IAM user and per-answer cost logging.
  • Architected a two-tier, offline-first platform for maritime environments: a cloud instance serving 10,000+ users plus Docker-based edge instances running fully offline on vessels with periodic sync. Built the Python sync agent (5-minute cycle, offline queue, exponential backoff, FK-safe ID mapping) and the FastAPI cloud sync API with API-key auth.
  • Led the Moodle 3.10 to 4.5 LTS (IOMAD) upgrade: rehearsed the cutover in Docker on production data, fixed a collation mismatch that would have aborted it, and wrote the cutover and UAT runbooks. Upgraded RDS MySQL 8.0 to 8.4 via Blue/Green.
  • Own the full image lifecycle for 500+ Linux hosts at remote sites: LUKS-encrypted data partitions, local RPM repository, anti-tampering checks, automated TLS renewal and provisioning bootstrap. Unattended hosts stay patched, encrypted and trusted with no engineer on site.
  • Maintain golden Windows and Linux AMIs with Packer (preinstalled applications, drivers, GPOs). The production Moodle AMI is baked from GitHub Actions with OIDC role assumption, so no AWS keys are stored in CI.
  • Built a Node.js tool (Prisma, AdminJS, S3) that prepares the course catalogue for import into Adobe Learning Manager: content upload, S3 reconciliation and import-ready CSVs.
  • Automated recurring operational work in Python and Bash: course packaging into GPG2-signed encrypted artifacts pushed to S3, bulk user import and log analysis.
  • Handle 2nd/3rd-line incident response across infrastructure and applications, including root-cause analysis and the follow-up work that prevents repeats. Maintain the team's technical documentation.
  • Own the LMS application lifecycle: version upgrades, performance tuning, security hardening, proctoring, and custom auth, reporting and user-sync APIs.
  • AWS
  • Terraform
  • Ansible
  • Packer
  • Docker
  • GitHub Actions
  • GitLab CI/CD
  • Jenkins
  • Python/FastAPI
  • Node.js
  • Bash
  • Amazon Bedrock
  • Redis
  • Nginx
  • MariaDB
  • Grafana
  • Linux
  • Windows Server

IT Manager, Aldar Properties PJSC

to

Astana. Abu Dhabi Plaza, a 500,000 sq m mixed-use development (offices, residential, shopping mall). End-to-end IT ownership.

  • Led the in-house NOC and helpdesk team, setting service standards and on-call rotations that gave every incident a defined owner and escalation path around the clock.
  • Delivered new systems from design through commissioning to operational handover, then ran them in production.
  • Operated the estate end to end: HPE LAN (100+ access, 10 distribution, 2 core switches), Fortinet and Palo Alto firewalls, Cisco CUCM IP-PBX (6 nodes), Aruba Wi-Fi (120+ APs), 10 HPE ESXi hosts running 30+ VMs, Bosch CCTV (3,000+ cameras), BMS and access control.
  • Managed the IT budget, vendor selection, procurement and compliance, holding suppliers to delivery commitments on a fixed-deadline construction program.

Head of IT Department, IEC Telecom Group

to

Almaty. Regional satellite communications operator (VSAT and L-Band) covering the Caspian region.

  • Led the IT department and directed VSAT and L-Band deployments across the region, standardizing how projects were built and handed over to support.
  • Designed mobile satellite IP, voice and video-streaming solutions for enterprise customers operating outside terrestrial coverage.
  • Migrated Active Directory and mail to Microsoft cloud services and a Windows Server 2012 R2 domain, retiring ageing on-premises dependencies.
  • Owned 1st to 3rd-line support for customers and staff across servers, PBX, CCTV, WAN/LAN, billing systems and databases.

Technical Support Engineer, IEC Telecom Group

to

Almaty. Satellite-phone support (Thuraya, Iridium, Inmarsat), Windows Server 2003/SBS and CentOS 5 administration. Built an internal billing system on Yii and PostgreSQL to replace manual reconciliation.

Systems Administrator, North Kazakhstan Energy Centre

to

Windows Server 2003 estate for 500+ users. Drafted and rolled out the organization's IT security policy.

Support Engineer, Halyk Bank of Kazakhstan

to

Banking software maintenance and corporate internet-banking support in a regulated financial environment. Network and PBX administration.