Yevhen Nezhyvyi

I build the internal tools a company's operations run on.

Claude Certified Developer — Foundations Anthropic · Sep 2026 · verify ↗

I ran B2B campaign operations for eight years, five of them as a team lead. Since late 2025 I have built the internal tools that run that work: seven systems, each built alone; four of them have run in production.

Email me

Looking for a role next to the COO or CEO: owning internal tools and operational automation. Full-time or contract. Based in Germany, remote across the EU.

What I do

  • Find the manual work

    I start with the people who do the work by hand. I write down each step, how long it takes and where it breaks.

  • Build the tool

    Spec, data model and the screen people work in. A model only where it helps, and plain rules where the answer has to be exact.

  • Keep it running

    I stay with the tool after launch. Health checks, cost tracking for model calls, and a clear "unavailable" instead of a silent failure. The team that did the work by hand runs the tool afterwards.

Built

Inside a 1,000-person B2B demand-generation company, 2025–2026. Described by what they do; the names and the data stay with the company.

Quality platform

In production
BeforeQA issues, client returns, audits and delivery pace sat in spreadsheets, email threads and the CRM.
NowOne platform with 57+ user accounts. It tracks issues and returns, assigns audits and verification reviews to people, and flags campaigns falling behind pace across about 1,200 campaigns. Five other internal tools read from it.

Client briefs

In production
BeforeEvery client brief was read and turned into targeting by hand.
NowRules read the brief first. Then a model drafts the targeting and a second model call audits it. Every label must exist in the company taxonomy; anything it cannot place is flagged for a person, not dropped.

Account lists and company research

Staging
BeforeClient account lists arrived in many layouts and were cleaned by hand.
NowThey are merged with the source of every row kept, and companies are matched to domains step by step. The rest is researched on the web, and the model may answer only from pages it actually found. Unclear ones go to a person instead of being guessed.

Content checks

In production

Before a campaign runs, the tool reads the client's asset (PDF, web page, Drive file or video) and checks it against the targeting. It flags a misfit with a quote from the asset, and the manager decides.

Design requests

In development

A Teams bot that turns a free-text request to the design team into a brief and returns the finished design without email. If the model fails, the request still goes through as the person wrote it.

Email nurturing

Handed over

The first version of the company's AI email-nurturing platform, built alone between December 2025 and January 2026: each email is written from the campaign's asset. Handed to the development team in January 2026.

Email infrastructure

Side project · handed over

Sending setup for a managed email service, 2025–2026: dedicated IPs with warmup, SPF, DKIM and DMARC, and blacklist and inbox-placement monitoring.

How I work
  • For each step I decide with the team whether it gets automation, a model or stays with a person.
  • When a tool is unsure, it asks. It does not guess.
  • Every change passes automated tests before it reaches production, and in most projects an automated review as well.
Background
2026
Software Developer, Internal AI Tools
2025
AI BuilderBuilt the department's internal tools alongside my project coordinator role.
2024
Project CoordinatorDepartment KPIs, workload across coordinators, technical requests.
2019
Team LeadA team of up to 10 running B2B lead-generation campaigns: checks before delivery, returns, data-quality standards.
2017
Campaign CoordinatorCampaigns end to end, reporting in Excel and Google Sheets.
2013
Customer support and serviceCall centre and e-commerce roles in Kharkiv, Ukraine.

2017 to now at one 1,000-person B2B demand-generation company, remote. Master's in Marketing, Ukrainian State University of Railway Transport. English (professional), German (basic), Ukrainian, Russian.

Stack

TypeScriptPythonNestJSFastAPIReactPostgreSQLRedisBullMQClaude APIClaude Coden8nGoogle Apps ScriptGitHub ActionsDocker

Contact
evgeniy.nezhivoy@gmail.com Dorsten, Germany