Zlatko Alomerović

Technical consultant · Ruby on Rails · AI workflow engineering

zlatkocodes.com · linkedin.com/in/zlatkoalomerovic · github.com/zalom · (EU citizenship)

Resume (PDF download)

Technical consultant and software engineer since 2015. I diagnose delivery problems, advise engineering leaders, and coach teams to deliver and own their work. I stay hands-on, turning technical decisions into working software. I engineered and built Plastic, an AgentOS for agent orchestration and graph-based engineering workflows, with separate implementation and verification agents.

AgentOS for agent orchestration and graph-based engineering workflows

AI engineering and research

  • Context engineering: Codified software delivery as dependency graphs with deterministic controls for task readiness, verification, retries, and acceptance. Used bounded task inputs, recorded decisions, savepoints, and handoffs to preserve context across agent sessions.
  • Orchestrator-workers and advisor: Coordinated task-specific agents and added consultation with stronger models for difficult engineering decisions. Workers returned structured results; the working agent retained decision ownership.
  • Evaluator-optimizer: Separated implementation and verification agents. Used verification findings to drive bounded correction and re-verification before accepting changes.

AI workflows for LinkedIn writing and job search

  • Independent output evals: Built an AI-assisted LinkedIn writing tool with deterministic checks and LLM-as-judge grading for factuality, writing structure, and instruction following.
  • Prompt chaining and parallelization: Connected posting triage, parallel evidence gathering, fit assessment, and decision support in a job-search assistant. Preserved intermediate results for review.
  • Typed judgments: Compared career evidence with job requirements using Jev. Retained source references and model and prompt versions so each assessment could be inspected and challenged.

AI infra and AI product discovery

  • Experimenting locally with vLLM and learning to route agent requests through LiteLLM to local inference and NVIDIA NIM models.
  • Used AI-assisted desk research to compare 72 service ideas through buyer evidence, spend, and ROI assumptions. Planning an open-source suite combining career evidence, job assessment, and AI-assisted application workflows.

Education

MEng. in Electrical Engineering and Computer Science

Graduate thesis: Application of the basic numerical methods in the dynamic systems simulations

Consulting and engineering experience

Cape Ann Enterprises: Senior Technical Consultant

Nov 2025 - Jul 2026 · Client: Learning Pool, WorkRamp division

  • Investigated the application and assessed team capabilities, then agreed staged onboarding with Learning Pool’s Director of Engineering. The team contributed around month two and handled greater complexity across two products and support in month three, earning commendations.
  • Coached five engineers through individual one-to-ones, Ruby training, and AI-assisted delivery practices, aligned with the client’s expectations. Introduced smaller, explainable pull requests and ongoing review. Supported one engineer’s progression from frontend to full-stack development with Ruby on Rails and React.
  • Anticipated the workload increase around month four and helped replace a front-end position with QA automation within existing headcount. The engineer joined and became a strong contributor.
  • Coached two Cape Ann engineers toward engineering management and staff engineering responsibilities, based on their client-facing and technical strengths.

Recrubo / Carv: Senior Software Engineer

Nov 2024 - Nov 2025

  • Managed GCP infrastructure for a year: VM deployments for background jobs, application deployments through Cloud Run, monitoring and fixes, IAM management, and security policy reviews. Led a teammate’s implementation of environment-specific job metrics for autoscaling.
  • Introduced label-triggered CI to avoid unnecessary test runs and conserve GitHub Actions minutes. Migrated scheduled integration jobs to Clockwork and added webhook logging.
  • Built the Bullhorn integration across authentication, API resources, event processing, and an embedded inbox. Verified existing AI workflows with integrations.
  • Resolved penetration-test findings across the backend API, web application, GCP, and Cloudflare, achieving a grade A pass.

Mahalo / Arubani Labs: Senior Technical Consultant

Oct 2023 - Feb 2024 · Part-time

  • Stabilized a live donation platform and its release process. Implemented staff invitations and role-based account access, fixed authorization checks, and added support impersonation and background email processing.

Marketer Technologies: Senior then Lead Software Developer

Apr 2021 - Apr 2023 · Lead from Mar 2022

  • Led distributed teams through the M2-to-M360 transition, managing four engineers and later three additional engineers on the legacy application. Maintained code review and ownership.
  • Planned and executed deliveries across distributed systems in coordination with the VP of Engineering and CTO. Built property-video delivery, CRM lead delivery, and advertising-platform integrations.

Ideaology: Technical Consultant

Apr 2021 - Mar 2022 · Concurrent contract

  • Guided technology-stack decisions and development direction, and led technical recruitment.

Coflyt: Senior Software Engineer

Jun 2020 - Feb 2021

  • Automated aircraft airworthiness-directive imports to replace manual entry. Moved external calls into background workers with timeouts and added browser tests and static security scanning.

Up and Running Software: Software Engineer

Jul 2017 - Jun 2020

  • Built dashboards, reporting, refunds, and chargeback handling for Passare. Took responsibility for shared release work and later built a new product’s Vue frontend against a mock API.

Earlier engagements

  • SABE Online · Apr 2017 - Jul 2017: Delivered content-management APIs and screens; corrected cross-company data exposure through scoped access.
  • Codeplace.com · Nov 2016 - Mar 2017: Built a GitLab-to-course publishing workflow and Rails teaching applications.
  • MiaVision · May 2016 - Dec 2016: Built Ember interfaces for three client products, including an interactive property floor-plan tool, with acceptance tests.
  • Dalia d.o.o · Mar 2015 - Dec 2015: Built an internal C# and SQL Server invoicing application and contributed to a team-built HR application.

Technical foundation

  • Stack-agnostic in approach, with deepest experience in my preferred stack: Ruby on Rails, PostgreSQL, and JavaScript.
  • API integrations · Distributed systems · Background processing · Observability · Application security · Unit testing · Mutation testing · Acceptance testing (with Varar) · CRAP · CI/CD · GCP / Cloud Run / Compute Engine · AWS / EC2
  • Also worked with Vue, React, Hotwire, C# / .NET, SQL Server, Ember, and AngularJS.
  • Rails and Ruby. Monoliths from Rails 4.1 to Rails 8, API-only backends, Hotwire, Sidekiq, PostgreSQL, RSpec, Minitest, Capybara, Playwright.
  • Integrations. About 35 external systems across ATSs (applicant tracking systems), payment providers, ad platforms, CRMs, government data feeds, and AI services; OAuth, webhooks, retry and event-processor patterns.

Highlights

2024

2019

2017

2016

  • Mentor and lecturer at “Europe CodeWeek 2016 - Peace of Code”, Tuzla, 20–23 October
  • Mentor and lecturer at “Junior Ruby on Rails Web Developer Course”, Tuzla, June 20
  • Mentor and lecturer at “freeCodeCamp Tuzla Workshop”, Tuzla, May 9
  • FreeCodeCamp campsite leader

2010

  • Lecturer on behalf of Leftor doo. at “GIS and GPS Workshop”, EESTEC and IEEE event, Tuzla

2009

  • Co-Founder and Contact Person, EESTEC, JLC Tuzla, Tuzla